Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
TRANSACTIONS
OF
J W R I
Vol. 39 No. 2
2010
OSAKA UNIVERSITY
JAPAN
Organization and Staff (December 2010)
Director General Dr. NAKATA Kazuhiro
1. Division of Materials Processing System
1) Energy Control of Processing
Professor Dr. TANAKA Manabu
Assistant Professor Dr. TASHIRO Shinichi
2) Energy Transfer Dynamics
Professor Dr. SETSUHARA Yuichi
Assistant Professor Dr. TAKENAKA Kousuke
3) Energy Processing Systems
Professor Dr. NAKATA Kazuhiro
Professor(Supplementary assignment) Dr. TAKAHASHI Yasuo
Assistant Professor Dr. TSUMURA Takuya
Assistant Professor(Supplementary assignment) Dr. MAEDA Masakatsu
Specially Appointed Researcher Dr. ZHANG Di
Specially Appointed Researcher Dr. LIU Hong
Specially Appointed Researcher Dr. LIU Duo
4) Advanced Engineering Science
Guest Professor Dr. NISHIYAMA Hideya
2. Division of Materials Joining Mechanism
1) Welding Mechanism
Associate Professor Dr. SHIBAYANAGI Toshiya
Associate Professor Dr. TAKAHASHI Makoto
Specially Appointed Researcher Dr. AHMED KHODIR Saad
2) Laser Welding and Materials Processing
Professor Dr. KATAYAMA Seiji
Associate Professor Dr. KAWAHITO Yousuke
Specially Appointed Researcher Mr. YOSHIDA Satoru
3) Composite Materials Processing
Professor Dr. KONDOH Katsuyoshi
Visiting Professor Dr. UMEDA Takateru
Assistant Professor Dr. UMEDA Junko
Specially Appointed Researcher Dr. IMAI Hisashi
Specially Appointed Researcher Mr. MURAKI Yoshinori
Specially Appointed Researcher Dr. LI Shufeng
3. Division of Functional Assessment
1) Mathematical Modeling and Computational Analysis
Professor Dr. MURAKAWA Hidekazu
Associate Professor Dr. NAKACHO Keiji
Associate Professor Dr. SERIZAWA Hisashi
Specially Appointed Professor Dr. RASHED Sherif
Specially Appointed Associate Professor Mr. YAMAMOTO Takao
Specially Appointed Researcher Dr. KAWAHARA Atsushi
Specially Appointed Researcher Dr. YAMAZATO Kunihiko
Specially Appointed Researcher Dr. MORISHITA Tashika
2) Dependability and Optimum Design
Professor Dr. KIM You-Chul
Assistant Professor Dr. SAKINO Yoshihiro
Specially Appointed Assistant Professor Dr. HIROHATA Mikihito
3) Materials Diagnosis and Life Assessment
Professor Dr. FUJII Hidetoshi
Specially Appointed Associate Professor Dr. SUN Yufeng
4. Smart Processing Research Center
Director Dr. NAITO Makio
1) Smart Beam Processing
Associate Professor Dr. ABE Nobuyuki
Associate Professor (Lecturer) Dr. TSUKAMOTO Masahiro
2) Smart Coating Processing
Professor Dr. NAITO Makio
Associate Professor Dr. ABE Hiroya
Specially Appointed Associate Professor Dr. OKUMIYA Masataro
Specially Appointed Researcher Dr. KONDO Akira
Specially Appointed Researcher Dr. YAMANAKA Shinya
3) 3D Nano/Micro Structure Control
Associate Professor Dr. KIRIHARA Sousyu
Specially Appointed Researcher Dr. TASAKI Satoko
4) Nano-Sized Particle bonding
Professor Dr. KOMIZO Yuichi
Associate Professor (Lecturer) Dr. TERASAKI Hidenori
Specially Appointed Researcher Dr. ZHANG Xinfang
5) Smart Green Processing
Associate Professor Dr. NISHIKAWA Hiroshi
5. Toyo Tanso “Advanced Carbon Design” Joint Research Chair
Guest Professor Dr. TOJO Tetsuro
Specially Appointed Associate Professor Dr. CHEN Weiwu
Specially Appointed Associate Professor Dr. NAKAMURA Fumishige
Guest Academic Staff Mr. NAKAMURA Masaharu
Guest Researcher Dr. MIYAMOTO Yoshinari
Guest Researcher Dr. MATSUMOTO Taihei
6. Fuji Electric Power Device Smart Joining Joint Research Chair
Guest Professor Dr. UMIDA Hidetoshi
Specially Appointed Professor Dr. TAKEMOTO Tadashi
Associate Professor Dr. NISHIKAWA Hiroshi
Specially Appointed Assistant Professor Mr. SHIMODA Masayoshi
Guest Academic Staff Mr. SHIOKAWA Kunio
Guest Academic Staff Dr. HIDAKA Noboru
Guest Researcher Mr. WATANABE Hirohiko
7. Development Base of Advanced Materials Development and Integration of Novel Structured Metallic and Inorganic Materials
Associate Professor Dr. KOBAYASHI Akira
Specially Appointed Associate Professor Dr. OHARA Satoshi
Specially Appointed Assistant Professor Dr. TERAJIMA Takeshi
Specially Appointed Researcher Dr. TAN Zhenquan
8. Global Collaborative Research Center for Computational Welding Science (CCWS)
Professor Dr. MURAKAWA Hidekazu
Specially Appointed Professor Dr. OKUMOTO Yasuhisa
Guest Professor Dr. TERASAKI Toshio
Associate Professor Dr. NAKACHO Keiji
Associate Professor Dr. SERIZAWA Hisashi
Guest Associate Professor Dr. MA Ninshu
9. Joint Interface Microstructure Characterization Room
Associate Professor Dr. TAKAHASHI Makoto
ISSN 0387-4508
TRANSACTIONS
OF
J W R I
Vol. 39 No. 2
2010
OSAKA UNIVERSITY
JAPAN
Transactions of JWRI, Vol. 39 No. 2, 2010
CONTENTS
Numerical visualization on melting and solidification of micron-sized metallic particles by laser irradiation
TAKASE Kazuyuki, MURAMATSU Toshiharu and SHOBU Takahisa ....................................1
Weld pool development and stirring behavior of AlCu4SiMg aluminum alloy during GTA welding hybrid
a longitudinal electromagnetic field
LUO Jian ....................................7
Numerical simulation of molten pool flow for various welding parameters in V-groove GMA pipe welding
CHO Dae-Won, NA Suck-Joo, CHO Min-Hyun and LEE Jong-Sub ....................................9
Numerical simulation and vision-based sensing of key-holing process in plasma arc welding
LIU Zuming, WANG Xiaojie and WU Chuan Song ..................................14
Experimental and numerical study of friction stir welding with double-shaft stir probe
XU Zhongfeng and LU Hao ..................................22
Experimental and numerical studies of material flow during welding by friction stirring
SHIMODA Yoichiro, TSUBAKI Masami, YASUI Toshiaki and FUKUMOTO Masahiro ..................................25
Effects of tool geometry and process conditions on material flow and strength of friction stir spot welded joints
HORIE Shohei, SHINOZAKI Kenjin, YAMAMOTO Motomichi,
KADOI Kota, NAKASHIN Hiroki and NORTH T. H. ..................................28
Numerical simulation of laser welding processes with CIP finite volume method
YAMASHITA Susumu, YONEMOTO Yukihiro, YAMADA Tomonori,
KUNUGI Tomoaki and MURAMATSU Toshiharu ..................................37
Modelling of the electron beam welding process applied to aircraft engine components
LAURENT D’Alvise ..................................40
Visualization of fluid flow and heat transfer in resistance spot weld nugget
LI Yongbing, SHEN Qi and CHEN Guanlong ..................................42
Development of welding method for wide gap lap joint of steel sheet using laser welding with hot-wire
YAMAMOTO Motomichi, SHINOZAKI Kenji, KADOI Kota, FUJITA Daigo,
INOUE Takeshi, FUKAHORI Mitsugu and KITAHARA Yoichiro ..................................44
Characteristics and mechanism on the distortion of friction stir welded aluminum alloy sheet
YAN Dong-yang, WU Ai-ping, SILVANUS Juergen, ZHANG Zeng-lei and SHI Qing-yu ..................................56
The influence of the solid state phase transformation on welding deformation of low alloy high strength steel
YANG Xiao, CHEN Junmei and LU Hao ..................................59
Residual stress measurement of large-bore stainless steel pipe with butt-welded joint by inherent strain method
MAEKAWA Akira, NAKACHO Keiji, MA Ninshu and SATO Reiko ..................................65
Structure optimization of air conditional compressor based on reverse welding deformation model
LI Yongzhi, LU Hao, CHEN Junmei and REN Liping ..................................71
Study on prediction of welding deformation for large-scale structure by T-E-P FEM using 3D shell element
ZHANG Xueyuan, LUO Yu and WANG Yang ..................................73
Fast finite element stress and deformation prediction for large thick-wall welded cylinder with angle-inserting elbow
ZHANG Kerong, ZHANG Jianxun, HUANG Siluo and QIU Yiqiang ..................................76
Thermal elastic plastic analysis for welding problem of large scale models
TANAKA Norihiro, KAWAHARA Atsushi,
SERIZAWA Hisashi and MURAKAWA Hidekazu ..................................79
Prediction of welding deformation for double bottom structure in cargo hold of 50000 DWT multipurpose ship
LI Jing and LUO Yu ..................................94
Flowing behaviors affected by different parameters and multi-materials in GTA weld pool hybrid
a longitudinal electromagnetic field
LUO Jian ................................100
Concept of inherent strain, inherent stress, inherent deformation and inherent force for prediction of
welding distortion and residual stress
MURAKAWA Hidekazu, DENG Dean and MA Ninshu ................................103
Inherent strain calculation from inverse analysis of measured welding deformation based on python of
ABAQUES CAE
ZHAO Haiyan, NIU Wenchong, WANG Peng,
YU Xingzhe, HE Hongwen and SUGIMURA Tadashi ................................106
ii
Application of inherent deformation and interface element to prediction of welding distortion during
assembly process
DENG Dean, MURAKAWA Hidekazu, MA Ninshu and SERIZAWA Hisashi ................................109
In-situ Observation of martensite transformation and retained austenite in supermartensitic stainless steel
ZHANG Shuoyuan, TERASAKI Hidenori and KOMIZO Yu-ichi ................................ 115
Prediction of hardness in HAZ of low-alloy steel produced by temper bead welding using neural network
YU Lina, NAKABAYASHI Yuma, ITO Shinsuke, KAMEYAMA Masashi,
HIRANO Shinro, CHIGUSA Naoki, SAIDA Kazuyoshi,
MOCHIDUKI Masahito and NISHIMOTO Kazutoshi ................................127
Surface tension of molten iron measured by oscillating droplet method using electromagnetic levitation
- Influence of oxygen adsorption on surface tension -
OZAWA Shumpei, TAKAHASHI Suguru, SUZUKI Shoji,
SUGAWARA Hiroharu, HIBIYA Taketoshi and FUKUYAMA Hiroyuki ................................130
Effect of welding sequences on the microstructure of electron beam welded TA15 titanium alloy and
304 stainless steel joints with copper filller metal
WANG Ting, ZHANG Binggang, CHEN Guoqing and FENG Jicai ................................133
Effect of grain size on solidification cracking susceptibility of type 347 stainless steel during laser welding
SHINOZAKI Kenji, WEN Peng, YAMAMOTO Motomichi,
KADOI Kota, KOHNO Yusuke and KOMORI Takuo ................................136
Image processing analysis of growth of carbide particles promoting SR embrittlement in HAZ of 2 1/4Cr-1Mo steel
KAWAKAMI Hiroshi, TAMAKI Koreaki, SUZUKI Jippei,
TAKAHASI Kanta, IMAE Yousuke and OGUSU Soichiro ................................139
Numerical simulation of impact toughness of welded joints for X80 pipeline steel
ZHANG Jianxun, XIONG Qinren, YANG Zhongna and ZHENG Li ................................141
Numerical analysis of deformation and thermal behavior during ultrasonic Al ribbon bonding
SUZUKI Shinji, OYAMA Yusuke, MAEDA Masakatsu and TAKAHASHI Yasuo ................................143
Visualization of heat and stress flows in thermodynamic crystals fabricated by laser scanning stereolithography
KIRIHARA Soshu, UEHARA Yasunori, TAKINAMI Yohei and TASAKI Satoko ................................145
Visualizations of terahertz frequency amplifications in water cells introduced into alumina diamond photonic crystals
OHTA Noritoshi, NIKI Toshiki and KIRIHARA Soshu ................................151
Nonlinear ultrasound and its applications in quality inspection and damage assessment in metallic materials
JIN-YEON Kim, LAURENCE Jacobs and JIANMIN Qu ................................157
iii
Measurement of fatigue damage parameter by sacrificial test piece and thermography
SAKINO Yoshihiro, SAKAGAMI Takahide and KIM You-Chul ................................159
Role of ambient pressure on splat formation and coating adhesion strength during thermal spraying process
YANG Kun, EBISUNO Yoshinobu, TANAKA Kazuhiro, USAMI Takashi,
FUKUMOTO Masahiro, YASUI Toshiaki and YAMADA Motohiro ................................162
Development of wettability evaluation equipment for solder paste using laser displacement method
KOYAMA Shinji, OYA Issei, ISAKA Toshihiro, SHOHJI Ikuo,
NISHIMURO Masashi, HIRAMOTO Kiyoshi and MIYAMOTO Hironaga ................................169
Examination of improvement effect of surface modification of Cu with organic acid on solder paste wettability
using a laser displacement meter
KOYAMA Shinji, AOKI Yukinari and SHOHJI Ikuo ................................172
Effect of welding direction on weld bead formation in high power fiber laser and MAG arc hybrid welding
MURAKAMI Takahiro, SHIN Min-Hyo and NAKATA Kazuhiro ................................175
Numerical analysis on effects of power source characteristics on arc properties in gas tungsten arc
TSUJIMURA Yoshihiro, TASHIRO Shinichi and TANAKA Manabu ................................178
Experimental observation of cleaning action of cathode spots in AC TIG welding of aluminum plates
TASHIRO Shinichi, SAWATO Hiroshi and TANAKA Manabu ................................180
Numerical simulation of heat source properties of pulsed tungsten inert gas arc
ITO Kuniyoshi, TASHIRO Shinichi and TANAKA Manabu ................................182
Numerical analysis of weld pool formation mechanism in TIG welding in consideration of the influence of
emitter material adding to the tungsten cathode
ZENIYA Tasuku . TASHIRO Shinichi, TANAKA Manabu,
YAMAMOTO Eri, YAMAZAKI Kei and SUZUKI Keiichi ................................184
Experiment and numerical simulation in temperature distribution and welding distortion in GMA welding
YAMANE Satoshi, YAMAZAKI Takuya, KANETA Tomoaki,
NAKAJIMA Toru and YAMAMOTO Hikaru ................................190
Improvement of bead formation in plasma MIG welding process in pure argon atmosphere
KATAYAMA Tsubasa, TASHIRO Shinichi and TANAKA Manabu ................................195
iv
Fundamental research on micro discharge process
- Research on discharge of sub-millimeter size process -
MINGON Park, HIRATA Yoshinori and NOMURA Kazufumi ................................207
Numerical simulation of fusion zone shape of lotus-type porous metals produced by laser welding
TSUMURA Takuya, NAKAJIMA Hideo and NAKATA Kazuhiro ................................213
Numerical simulation on the coupled arc and pool for GTAW using a unified mathematical model
YONGPING Lei, YAOWU Shi, JIAN Lin,
ZHENYANG Lu, RONGSHI Xiao and H. Murakawa ................................215
Prediction of 475°C embrittlement in stainless steel welds using phase field model
SAIDA Kazuyoshi and NISHIMOTO Kazutoshi ................................218
Interfacial reaction between Sn-3.0Ag-0.5Cu solder / Co-P plating and Ni-Co-P plating
DAITO Tomoya, NISHIKAWA Hiroshi, TAKEMOTO Tadashi and MATSUNAMI Takashi ................................223
The dissymmetry of friction stir welding joints and variable polarity plasma arc welding joints study
CHEN Shujun, WANG Long and YU Yang ................................228
Industrial application of welding temperature field and distortion visualization using FEA
RAPHAEL Thater, WILLIAM Perret, CHRISTOPHER Schwenk,
UWE Alber and MICHAEL Rethmeier ................................231
Prospective design of weld joint between first and side walls in the test blanket module for ITER
NAKAMURA Shinichiro, SERIZAWA Hisashi,
TANIGAWA Hiroyasu and MURAKAWA Hidekazu ................................234
Preliminary numerical research of microstructural fracture behavior in metal by using interface element
TOMIYAMA Seigo, SERIZAWA Hisashi, HAJIMA Tsuyoshi and MURAKAWA Hidekazu ................................237
Visualization of mechanical properties in alumina dental crowns fabricated by using laser scanning stereolithography
TASAKI Satoko, KIRIHARA Soshu and SOHMURA Taiji ................................248
Combinatorial analysis of plasma-polymer interactions for formation of inorganic-soft materials hybrid structure
TAKENAKA Kosuke, CHO Ken, SETSUHARA Yuichi,
SHIRATANI Masaharu, SEKINE Makoto and HORI Masaru ................................250
Materials joining technologies and interface science for integration of novel structured metallic
and inorganic materials
SETSUHARA Yuichi and NAKATA Kazuhiro ................................259
Approach for the project on advanced materials development and integration of novel structured metallic
and inorganic materials by materials and structures laboratory of Tokyo Institute of Technology
OKADA Kiyoshi ................................264
Fabrication and thermoelectric properties of eco-friendly silicides for thermoelectric power generation
using waste heat
ITOH Takashi ................................276
Fabrication of zirconium oxide solid electrolytes with ordered porous structures by using micro stereolithography
KIRIHARA Soshu and TASAKI Satoko ................................279
Room-temperature Coulomb oscillation of Ni-Nb-Zr-H glassy alloys with nanoscale size clusters
M. Fukuhara, H.Yoshida and A.Inoue ................................289
Enhancement of solderability of Cu60Zr30Ti10 bulk metallic glass by dealloying in hydrofluoric acid solution
NAOI Takehiro, NISHIKAWA Hiroshi, TAKEMOTO Tadashi,
ABE Hiroya, FUKUHRA Mikio and INOUE Akihisa ................................293
Effects of photon irradiation in UV and VUV regions during plasma processing of organic materials
CHO Ken, TAKENAKA Kosuke, SETSUHARA Yuichi,
SHIRATANI Masaharu, SEKINE Makoto and HORI Masaru ................................298
vi
Development of a novel integrated iontophoresis electrode consisting of metallic and drug-loaded layers
T. Yoshioka, N. Saito, T. Ikoma, N. Ohashi, H. Haida,
S. Ando, R. Wakita, H. Fukayama, M. Umino and J. Tanaka ................................304
Developments of hybrid in-situ observation system to study the microstructural change of metallic alloys
TERASAKI Hidenori and KOMIZO Yu-ichi ................................316
Development of amorphous boron carbon oxynitride film for transmission electron microscope with
environmental-cell system
MATSUTANI Takaomi, TSUTSUI Hidenori and KAWAKAKI Tadahiro ................................322
Properties of metallic glass coatings sprayed by gas tunnel type plasma spraying
KOBAYASHI Akira, KURODA Toshio, KIMURA Hisamichi and INOUE Akihisa ................................325
Titanium oxide film deposition by vortex air thermal plasma assisted spray pyrolysis deposition
ANDO Yasutaka and KOBAYASHI Akira ................................335
Investigations on frequency shift probes for monitoring of electron conditions in nano-materials processing plasmas
NAKAMURA Keiji and SUGAI Hideo ................................338
Joining mechanism between aluminum and polypropylene resin using insert materials by laser irradiation
HINO Makoto, MITOOKA Yutaka and KATAYAMA Saiji ................................343
vii
Modeling of temperature distribution with metal vapour in pulsed TIG including influence of radiative absorption
IWAO Toru, MORI Yusuke, SAKAI Tadashi, TAKI Hiroyuki,
SHIMOKURA Takuya, TASHIRO Shinichi, TANAKA Manabu and YUMOTO Motoshige ................................346
Dynamic mechanical behavior of Sn-Ag-Cu lead-free solders by tensile test under high strain rate
YASUDA Kiyokazu, SAKINO Yoshihiro, SHOJI Ikuo and TAKEMOTO Tadashi ................................360
Oxide nanosheets and their assemblies for new ceramic joining and smart processing
OSADA Minoru and SASAKI Takayoshi ................................362
Evaluation of magnetite and related iron compounds in the teeth of chiton using X-ray and electron analyses
NUMAKO Chiya, SATO Kazuyoshi, ABE Hiroya and OHARA Satoshi ................................364
Characterication and control of nanoparticle dispersion behavior for smart processing in liquid phase
KAMIYA Hidehiro, IIJIMA Motoyuki, TAKENOUCHI Shun,
IINUMA Chihiro and NAITO Makio ................................366
Development of analytical method for welding mechanics using idealized explicit FEM
SHIBAHARA Masakazu and IKUSHIMA Kazuki ................................384
viii
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Numerical simulation) (Laser welding) (Thermal-hydraulics) (Phase change) (Metallic
particles) (Melting) (Solidification) (temperature) (liquid metal) (repair technology)
1. Introduction ª wD
P «
g
w D gU g , j º» qgw qgint * hgsat (5)
Aiming at the enhancement of safety of currently
operated nuclear reactors, development of a repair « wt wx j »
¬ ¼
technology using laser welding [1] has been performed. An
experimental study on welding of micron-sized metallic
particles on a stainless steel plate by laser irradiation has w
w D l Ul elU l , j
already been carried out. Moreover, an analytical study was
D l Ul el
wt wx j
also begun to predict numerically the experimental results.
The micron-sized metallic particle that is made of iron is
P « l
ª wD w D lU l , j º» q w qint * hsat (6)
heated by the laser irradiation and then melts and finally the l l l
« wt wx j »
solid metallic particle changes perfectly to the liquid metal. ¬ ¼
After a long time, the liquid metal solidified the solid metal
with decreasing the temperature. This paper describes 3. Analytical conditions
visualized numerical simulation results of the melting and
Figure 1 shows the analytical geometry, which represents
solidification behavior of the micron-sized metallic particles
the experimental condition. 12x12 metallic particles are set
that change from the solid to liquid and the liquid to solid.
into a slit on a metallic plate. Each of them is made of iron.
The outer diameter of the metallic particle is 0.04 mm. The
2. Governing equations
size of the slit is 0.48 mm square and 0.05 mm in depth. It
Governing equations consist of gas and liquid phases as briefly simulates any cracks on vessels or components of
follows: nuclear reactors. A laser beam is irradiated from the upper
-Mass conservation; side into the slit. The shape of the laser beam is round as can
w w be seen using red in Fig. 1, and the diameter is 0.48 mm.
wt
D g Ug
wx j
D g U gU g , j *
(1)
w w
wt
D l U l
wx j
D l UlU l , j * (2) I 0.48 mm
Heat flux Metallic particle
-Momentum conservation;
wU g ,i wU g ,i int
1 wP M g ,i
Ug,j 0.05 mm
wt wx j U g wxi D g U g Metallic 5 mm
* wW g ,ij 0.48 mm plate
D g Ug
U g ,i U l ,i D 1
gi (3)
g Ug wx j
Fig. 1 Analytical geometry.
wU l ,i wU l ,i 1 wP M l ,i
int
Ul , j
wt wx j Ul wxi D l Ul Figure 2 shows the phase change model from the solid to
liquid phase and conversely the liquid to solid phase. The
* wW l ,ij
D l Ul
U g ,i U l ,i D 1U wx j
gi (4) phase change behavior of the metallic particles with the laser
irradiation is obtained from the liquid fraction, L f , which is
l l
-Internal energy conservation; specified as the following equation:
1
Numerical visualization on melting and solidification of micron-sized metallic particles by laser irradiation
T TS Temperature (K)
Lf (7) 㻌 㻌 㻌 㻌 㻌 500 2200
TL TS
Here, T represents the temperature of the metallic particle, T S
solidification temperature, and T L melting temperature. The
following three states related to T and L f are considered:
i) Solid state (at L f =0); when T<T S ;
ii) Liquid state (at L f =1); when T>T S ; and,
iii) Phase change state; T S <T<T L .
Therefore, change of enthalpy in the material from the
solid to liquid by phase change is calculated from
T
h h0 ³ c p dT H L L f (8) (a) After 0.1 ms
T0
䠍
Liquid fraction Lf
Lf
Liquid
㻜
(c) After 0.4 ms
Solid
TS T TL
Temperature T
Fig. 2 Phase change model between solid and liquid phase.
2
Transactions of JWRI, Vol.39 (2010), No. 2
D : Volume ratio
* ,* +,*-: Gas generation rate [kg/m3/s]
Qt : Eddy viscosity [m2/s]
U : Density [kg/m3]
W : Shear stress tensor [kg/m/s2]
Subscripts
g : Gas
l : Liquid
i, j: Spatial coordinate component
References
[1] S. K. Maiti, et al.: Science&Technology of Welding and
Joining, 8 (2003), pp.377-384.
3
Transactions of JWRI, Vol.39 (2010), No. 2
†
Three-dimensional computational modelling of MIG welding
MURPHY Anthony㻌㻌 B*
KEY WORDS: (Computional modelling) (MIG welding) (Gas–metal arc welding) (Arc welding)
(Weldpool) (Droplets) (Arc Plasma) (Weld reinforcement)
4
Three-dimensional computational modelling of MIG welding†
5
Transactions of JWRI, Vol.39 (2010), No. 2
6
Transactions of JWRI, Vol.39 (2010), No. 2
LUO Jian*
KEY WORDS: (Welding pool) (GTAW) (Magnetic field) (Aluminum alloy) (Development) (Stirring motion)
1. Introduction
(a) (b)
In GTA Welding with a longitudinal electromagnetic
field (LMF-GTAW), the formation, microstructure and
mechanical properties of aluminum alloy are studied. t=0.02s t=0.04s
However, there is still a lack of practical or numerical
detailed understanding of the formation of the welding pool
during LMF-GTAW. For example, the effect of the
longitudinal electromagnetic field on the formation of the
weld pool is not clear yet. The additional electromagnetic
force can not only affect the behavior of welding arc, but
also change the force balance state and heat transfer of metal
molten body of the weld pool. (c) (d)
In this paper, we focus on the need for understanding the
weld pool’s formation is AlCu4SiMg aluminum alloy during
LMF-GTAW by a new mathematic model and finite element t=0.05s t=0.06s
method (FEM).
(g) (h)
t=0.2s t=0.3s
Fig.1 Diagram of FEM model in LMF-GTAW Fig.2 The gestation period of the LEM-GTA Welding pool
7
Weld pool development and stirring behavior of AlCu4SiMg
aluminum alloy during GTA welding hybrid a longitudinal electromagnetic field
(a) (b)
t=0.4 t=0.6
(c) (d)
8
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Numerical simulation) (Abel inversion) (Molten pool flow) (V-groove Pipe welding)
(Elliptic arc heat flux) (Bead shape)
9
Numerical simulation of molten pool flow for various welding parameters in V-groove GMA pipe welding †
5. Conclusion
Due to different welding positions, molten pool flows
(1) The elliptic symmetry arc heat flux model is
make different bead shapes. Figure 2 shows the molten
proposed for CFD pipe welding simulation
pool flow before solidification and fusion zone in vertical-
(2) For different welding position, molten pool flows and
down position. Figure 3 shows the result in overhead
fusion zones are different.
position.
Reference
[1] Y.T. Cho and S.J. Na: Meas. Sci. Technol. 16(2005),
pp.878-884.
[2] W.I. Cho, S.J. Na, M.H. Cho and J.B. Lee: Proc. LAMP
(2009), pp.251
[3] K.W. Park and S.J. Na: Applied Surface Science, 256
(2010), pp.2392-2399
[4] M. Ushio, D. Fan and M. Tanaka M.: Trans. JWRI. 22
(1993) pp.201–207
Fig.2 Molten pool flow and fusion zone [5] W. L. Wiese: Spectrochimica Acta, 46B (1991) pp. 831–
in vertical-down position 841
[6] C. Yubero, M.S. Dimitrijevic, M.C. Garcia M.C. and M.D.
Calzada: Spectrochimica Acta Part B, 62(2007), pp.169-
176,
[7] D.W. Cho, S.J. Na, M.H. Cho and J.S. Lee: 63rd IIW Doc.
(2010)
10
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (SPH) (Arc welding) (Weld pool) (Penetration shape) (Marangoni effect)
11
Three-dimensional simulation of a flow in an arc weld pool by SPH method
In Case 2, the fluid heated at the center on the surface approximation and is given to all liquid particles. As a
spreads outward, while in Case 3, the heated fluid is result, the high temperature region is observed only on the
transported downward. These results derive from the surface.
difference in the direction of the surface tension gradient. It Case 6 shows that the fluid heated at the surface flows
is known that Marangoni effect shows two different downward. During the arc welding, the electromagnetic
tendencies depending on the sulfur content in SUS304, i.e. field is generated, and consequently, the Lorentz force acts
the surface tension increases with temperature when the on the fluid inside of the weld pool. In this study, the
sulfur content is low, while it decreases when the sulfur axisymmetric Lorentz force is calculated from the electric
content is high [4]. In the present computation, the specific current density distribution obtained in Ref. [3].
data of Ref. [4] is used. Figure 2 depicts the computational results when all the
In Case 4, the gas drag, which is a frictional force by forces and the phase transition are taken into account, at t =
the shielding gas flow on the surface of the weld pool, is 20.0 s. For the phase change, the latent heat is treated. The
calculated from the velocity distribution obtained in Ref. red particles are at the melting point, 1750 K. The depth of
[3]. Because the drag acts outward to the surface particles, penetration is approximately 4 mm in Case 7, meanwhile
the high temperature region at the surface spreads. the penetration has reached about 8 mm in Case 8. This
In Case 5, the buoyancy caused by the non-uniform difference is obviously attributed to the difference in the
temperature distribution in a weld pool are taken into sulfur contents.
account. The force is calculated using the Boussinesq
12
Transactions of JWRI, Vol.39 (2010), No. 2
5. Conclusions References
In this study, the SPH method was applied to a [1] K. Yamazaki, E. Yamamoto, K. Suzuki, F. Koshiishi, S.
simulation of the thermofluid behavior during an arc welding Miyazako, S. Tashiro, M. Tanaka and K. Nakata: Quarterly
process, taking four dominant flow-driving forces into Journal of the Japan Welding Society, 27 1 (2009), pp.34-40.
(in Japanese)
account. The numerical calculations presented the effect of
each force on the temperature distribution inside the weld [2] M. Shigeta, T. Watanabe, S. Izawa and Y. Fukunishi: Int. J.
pool. Also, the simulation of a weld pool was conducted Emerging Multidisciplinary Fluid Sci., 1 (2009), pp.1-18.
including all the forces and the phase transition. The [3] M. Tanaka, M. Ushio, T. Isa and J.J. Lowke: JSME Int. J.
influences of the sulfur content on the flow behavior and the Series B, 48 (2005), pp.397-404.
weld penetration were shown. The penetration shapes of the
present computation showed a tendency similar to Ref. [3]. [4] T. Zacharia, S.A. David, J.M. Vitek and T. Debroy: Metall.
Trans. B., 21 B (1990), pp.600-603.
Here, we can conclude that SPH method is a method suitable
for a thermofluid simulation of an arc weld pool.
13
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Numerical simulation) (Sensing) (Plasma arc welding) (Key-holing process)
1. Introduction 180
160
2. Experimental measurement
The developed waveform-controlled PAW experimental
system consists of the computer, PAW machine, data
acquisition unit, welding current sensor, efflux plasma
voltage sensor and CCD camera. The computer is the
central unit for adjusting welding current, and sampling the Plasma
signals of welding current, efflux plasma voltage and arc
keyhole images. On one hand, a measuring bar mounted
underneath the workpiece to be welded and kept insulated
electrically is employed to detect the efflux plasma voltage
when the keyhole is established. On the other hand, a CCD
camera is aimed at the weld pool from the backside to
capture the keyhole images. Efflux
Figure 1 shows the measured welding current and plasma
efflux plasma voltage signals. At the dropping stage of
welding current from the peak level to the base level, two
sub-stages of current decreasing with different slopes are
added. The signal of the efflux plasma voltage is around Fig.2 The captured images of both the plasma arc and
zero before the keyhole is established, while it exceeds a the efflux plasma
certain value after the keyhole is established. It is clear that
each pulse produces one keyhole during the welding
process. Figure 2 is the images of both the plasma arc transient waveform-controlled PAW process are very
and the efflux plasma captured from a side view. Figure 3 complicated. As a first step, the level-set method is used to
shows an image of keyhole from the backside of the track the keyhole boundary in stationary continuous-current
workpiece. PAW process [3]. As shown in Fig. 4, the level-set function
is defined as a signed distance function M ( x , t ) r d ,
3. Simulation of keyhole shape and size where d is the actual distance from the keyhole boundary,
The modeling and simulation of keyhole behaviors in
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* Institute of Materials Joining, Shandong University, China Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
Japan
14
Numerical simulation and vision-based sensing of key-holing process in plasma arc welding
-2
(a)
z, mm
4
0.18 m/s
6
8
-8 -6 -4 -2 0 2 4 6 8
r, mm
-2
Fig.3 The captured image of keyhole at backside (b)
0
x is the space variables, and t is time. The level-set function
has the value of 0 on the keyhole boundary. The plus (minus) 2
sign denotes the outside (inside) of the keyhole boundary.
z, mm
By using the level-set method the keyhole boundary tracking
4
problem can be transformed into a partial differential
equation, which can be numerically solved with other 0.48 m/s
6
governing equations. The general form of the level-set
equation is
8
wM -8 -6 -4 -2 0 2 4 6 8
F M 0 (1)
wt r, mm
model is used to determine the evolution of both keyhole and 0.53 m/s
weld pool geometry at different time steps. Figure 5 6
Plasma 6
0.45 m/s
torch
8
-8 -6 -4 -2 0 2 4 6 8
ij<0 r, mm
16
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Pulsed arc) (High frequency) (Diffusing mode) (TIG welding) (Physical properties)
1. Introduction
Extensive research has been reported on pulsed arc
welding and its effects over the past decades. The
advantage of the pulsed arc summing-up in the literature,
however, mostly concerns weld metallurgy, including
refinement of fusion zone grain size and substructure,
reduced width of HAZ and so on [1]. Mechanical property
improvement of welding joint always was explained as an
evidence of the effects. To deepen the understanding of
pulsed arc physical properties, a study on the effects of
pulsed TIG welding arc is made from lower frequency to
higher frequency up to 80 kHz in both electric and thermal Fig. 2 Arc voltage vs pulse frequency
respects.
3. Micro View of Pulsed TIG Arc
2. Macro View of Pulsed TIG Arc Electron density in arc column is observed by Stark
Welding is carried out using a TIG welding machine broadening spectrum and can be calculated using Stark
(EWM Tetrix 500) as base current source (BCS). A high broadening of the 656.3nm HĮ emission line [3]. Figure 3
frequency current producing device is used as peak current shows a trend of the electron densities in the Ip period with
source (PCS), and it works with BCS synchronously for frequency changing. The electron density goes down at
generating square current waveform.Welding conditions higher frequencies though the input energy from power
are as follows ĭ0 mm, 2% Ce-tungsten electrode with source to arc is increasing. It may indicate that arc column
60º tip angle; Argon shielding gas 10l/min and 4 mm arc diffusing near the tungsten cathode instead of continually
length. Welding parameter setting᧶Ip = 100A, Ib = 50A. lift-up constriction at higher frequency in order to realise
Figure 1 shows the arc profiles changing in the “electric- energy balance.
thermal transfer” mode from a diffuse to a lift-up
attachment with pulse frequency increasing [2], and a ball-
like arc column contracting towards tungsten cathode can
be observed in the higher pulse frequency region. Figure 2
shows a trend of the arc voltage Up and Ub with frequency
changing respectively. The phenomena of Up going up with
the increase of pulse frequency and strong oscillation
appearing and taking 30% of the Ub period during the
changing from Ip to Ib are rather noticeable.
50 1K 2.5K 20K 50K 80K
4. Discussion
While energy frequency inputting to the arc near or
reaching to the order of the critical time of energy transfer
based on electron-collision mode[3], a positive potential
Fig. 1 A map of the arc profiles in different pulse drop would be needed in order to transport the energy from
frequancies
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* Beijing University of Technology, Beijing, China Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
College of Mechanical Engneering & Applied Electronics Japan
Technology
17
A Study on the macro-micro physical properties in pulsed arc plasma
5. Conclusions
Based on the above discussion, the following results
may be concluded:
(1) With the increase of pulse frequency, there are 2
diffusing modes in TIG arcs, that is, diffusing
attachment towards anode at lower frequency region
and ball-like diffusing near cathod at higher frequency
region. The phenomenon of continual constriction with
the frequency increasing is not as expectation, only a
Fig. 4 Arc dynamic impedance model slight constriction of the arc at the lower frequency
the arc column into the anode. It would be a reason to region can be identified.
explain the phenomena of Up going up with increase of (2) The electron density distribution with frequency
pulse frequencies. changing may support the profiles of pulsed arc in some
According to the experimental result, the TIG arc may diffusing modes. It may be helpful to understand the
be not the first-order inertial link under the higher physical properties of free burning TIG arcs.
frequency condition, but a second-order link with the
obvious oscillation phenomenon in the view of electricity. Acknowledgement
An arc dynamic impedance model may be shown in Fig. 4, This work is financially supported by the Chinese NSF,
and the corresponding arc dynamic impedance transfer project no. 50375005.
function may be expressed by:
References
[1] T. Senthil Kumar, V. Balasubramanian and M.Y. Sanavullah:
U (s) s Influences of pulsed current tungsten inert gas welding
Z (s) parameters on the tensile properties of AA 6061 aluminium
I (s) 1
Cs s
2 1 (1)
R L alloy, Materials and Design 28 (2007), pp. 2080–2092
[2] J Heberlein, J Mentel and E Pfender: The anode region of
According to the model, the inductance feature may electric arcs: a survey, J. Phys. D: Appl. Phys. 43 (2010),
play a main role at lower frequency region, and the pp.1-31
capacitance feature may play a main role at higher [3] Hans R. Griem: Spectral line broadening by plasmas. New
frequency region. It would be a reason to explain the York: Academic press, (1974), pp.226-231
phenomena of Ub going down.
18
Transactions of JWRI, Vol.39 (2010), No. 2
19
Evaporation phenomena of magnesium during pulsed-MIG arc welding of aluminum alloy
(mm 2 )
About the droplet growth behavior, when the droplet Sdrop
(mm 3)
Mg content, C Mg (mass %)
growth time TS is under 0.5 ms, both the droplet volume 6 5
drop
and droplet surface area increase slightly. From the high-
drop
speed video images, the arc root is seen to spread from the CMg
Dorplet volume, V
droplet itself deforms. Before the arc root spreads up to a 4 4
solid part of the welding wire, the droplet volume increased
only a little. When TS is over 0.5 ms to 1.6 ms, both the
droplet volume and the droplet surface area increase
abruptly. About the magnesium evaporation behavior, when Vdrop
2 3
TS is 0 ms, that is, at the end of the background period or
the beginning of the peak period, the magnesium content in
droplet is 4.63 mass %. With increasing the droplet growth
time TS up to 1.0 ms, the magnesium content decreased to
0 2
4.2 mass %. After that, the magnesium content decreased
very slowly to reach about 4.0 mass % at the end of the 0.0 0.5 1.0 1.5 2.0
peak period of about 1.6 ms. The decrease in magnesium Droplet growth time, TS (ms)
content with increasing TS is considered to be caused by Figure 2 Droplet volume, droplet surface area and
the evaporation of magnesium from the droplet surfaces magnesium content at different droplet growth times.
whose temperatures are reported to nearly reach their
boiling point [2]. From Fig. 2, empirical relationships
between the droplet volume V drop (mm3), droplet surface
area S drop (mm2), magnesium content C Mg (mass %) and TS After the above calculation, the evaporation rates V vap(i)
(ms) are obtained as Equation (1) to (3). (i=1, 2,䡡 䡡 㻘㻌䡡 㼚㻕㻌 and evaporation flux are plotted
accounting for TS. The calculated results are shown in Fig.
V drop =0.844+8TS2/(TS2+18) (1) 3. It is seen that in region I (TS=0-0.2 ms) slight decreases
S drop =3.405+8TS2/(TS2+3.2) (2) both in the evaporation rate and flux are confirmed. This is
C Mg =3.9+0.73e(-TS/0.85) (3) attributed mainly to the evaporation from the anode root
area. In region II, the evaporation rate increases very fast
From equation (1) to (3), the evaporation rate and flux of with increasing TS up to 1.2 ms and the evaporation flux
magnesium during peak period can be calculated as follows. increases slightly and finally reaches a saturated value.
Firstly, one peak period is divided by a time interval ԥTS When TS increases into region III, the increase of the
(=0.01 ms) into TS (i) (i=0, 1, 2, 㺃㺃㺃, n). V drop(i) and C Mg(i) evaporation rate with increasing TS slows down a little and
(i=0, 1, 2, 㺃㺃㺃, n) are calculated from Equation (1) and (2) the evaporation flux decreases slightly. The change from
by inputting TS=ixԥTS. Secondly, the evaporation rate region II to region III is considered to correspond with the
and flux are calculated as follows by putting i=1, 2, 㺃㺃㺃, n. formation of particulate fumes as shown in Figure 1. In
(1) Calculating the mass of magnesium W dMg(i-1) and region III the slow-down of the evaporation rate or the
W dMg(i) at TS (i-1) and TS (i) , respectively. decrease of evaporation flux with increasing TS is thought
(2) Calculating the mass of magnesium W wMg in the to be caused by the evaporation of aluminum accompanied
welding wire melted during ԥTS. with magnesium which is supposed to be included in
(3) Calculating the evaporation rate of magnesium V vap(i) particulate fumes.
[=(W dMg(i-1) +W WMg -W dMg(i) )/ԥTS].
(4) Calculating the evaporation flux by dividing V vap(i) by
Equation (2).
20
Transactions of JWRI, Vol.39 (2010), No. 2
5. Conclusions
4 1.0 This study is summarized as follows.
䊠 䊡 䊢 (1) According to the evaporation behavior of magnesium,
21
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (FSW) (Numerical simulation) (Heat source model) (Temperature field)
1. Introduction
Friction stir welding[1] (FSW) is a new developed
solid-state joining technique. FSW has been successful
applied to aerospace, ship manufacturing and automobile
industries. However, due to the disadvantage of the concave-down area
keyhole, which is caused by the probe’s pulling out when center line
circular weld is applied, NASA developed a kind of
double-shaft stir probe. The stir probe can pull out with a Fig. 2 Cross-section of the weld bead
certain velocity from the weldment when the shoulder is
still rotating and advancing along the weld bead. A
Significant process has recently been made on heat
schematic illustrating the process of FSW is shown in Fig.
transfer modeling for FSW. Frigaard, Grong, and
1.
Midling[2] developed a process model for FSW, the heat
input from the tool shoulder is assumed to be the friction
probe heat, and the coefficient of friction or the calculated
temperature during the welding is adjusted to keep the
calculated temperature from exceeding the material
melting point. The Rosenthal equation for modeling
shoulder heat-transfer for thin plates has also been applied in
modeling the heat-transfer in FSW[3]. Because of the
process of the probe’s pulling out, the heat transfer process
will become complicated. Numerical methods are very
convenient for the investigation of the mechanism of FSW.
Therefore, this work mainly studies the weld quality and
Pull
the heat transfer.
V
22
Experimental and numerical study of friction stir welding with double-shaft stir probe
23
Transactions of JWRI, Vol.39 (2010), No. 2
24
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Friction Stir Welding) (Material flow) (High speed video camera) (PIV) (FEM model)
25
Experimental and numerical studies of material flow during welding by friction stirring
Velocity (mm/s)
2.0
were used in this simulation: thermal conductivity k=0.17
(W/mK), and thermal capacity c=1.95 (N/(mm2K)). These RS
assumption linearizes the finite element formation for Flow2 1.0
temperature analysis and results in better convergence [5].
A rigid-viscoplastic temperature and strain rate dependent Rotation
material model is employed Work piece direction 0.0
()()
n
σ = cε ε + y
m
(1) Fig. 3 FEM analysis of material flow.
Where c =7.791, n =1.987, m =-0.116 and y =20.052
Figure 4 shows measured and calculated results of material
(MPa) are material constants determined by a numerical
flow velocities by PIV and FEM analysis during welding.
regression based on experimental data [6].
The material flow velocities of PIV analysis ranges from
about 2 to 20 mm/s at AS. The material flow velocities of
4. Results and Discussion
PIV analysis ranges from 1 to 5 mm/s at RS. These are
Figure 2 shows the material flow of the weld with high similar tendencies that of FEM analysis. Additionally, both
speed video camera and analysis of PIV at AS and RS. velocities of PIV and FEM at AS are faster than that flow at
Figure 3 shows FEM analysis of material flow. From PIV RS.
and FEM analysis, material flow was observed different One revolution of welding tool
type of flow direction at RS and RS. At RS, direction of the
material flow (Flow2) was corresponded to the tool rotation. 30
However, material flow (Flow1) at AS was in the reverse (a) AS PIV
25 FEM
Velocity, mm/s
(b) RS PIV
8 FEM
Velocity, mm/s
Pin
PVC Pin PIV
1mm (gray) 6 FEM
(c) PVC (gray) (d) 4
Pin Pin
2
Velocity (mm/s)
(d) 2.0
RS
Flow2
0
PVC 1.0 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
(clear) Time, s
Rotation Fig. 4 Measured and calculated results of material
Work piece direction 1mm 0.0 flow velocities by PIV and FEM analysis during welding.
Fig. 2 Material flow of weld with high speed video camera Analysis area is Fig. 2, (b), (d) and Fig. 3, (b), (c).
and analysis of PIV.
26
Transactions of JWRI, Vol.39 (2010), No. 2
5. Conclusions [2] Schmidt H et al, Material flow in butt friction stir welds in
Material flow by friction string during welding was AA2024-T3, Acta Materialia 54 (2006) 1199-1209.
investigated with experimental and numerical studies. [3] Y. Yasui et al, Macrostructure of butt weld between aluminum
Following results were obtained. and steel by friction stirring, Proc. 6th Int. Symp. on FSW
(2008) 1100-1130.
1. Material flow at RS corresponded to the rotating
[4] T. Shinoda et al, Observation of metal flow phenomenon in
direction of the tool. However, material flow at AS friction stir welding of aluminum alloy, Proc. 7th Int. Symp.
was in the reverse direction of the tool rotation. The on FSW (2008) 1405-1430.
material flow from AS to RS passed behind the pin. [5] G. Buffa et al, Finite element studies on friction stir welding
2. From PIV and FEM analysis, material flow velocities processes of tailored blanks, Computers and Structures 86
at AS are faster than that at RS. (2008) 181-189.
[6] Sharon Y.Soong et al, The effects of thermomechanical
References history and strain rate on antiplasticization of PVC, Polymer
[1] W. M. Thomas, International patent application, No. 49 (2008) 1440-1443.
PCT/GB92, GB Patent Application, No. 9125978.8.6, (1991).
27
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (FSSW) (Tool geometry) (Material Flow) (Strength) (Mg alloy)
1. Introduction since the stir zone is not grown up enough. However, when
FSSW has been employed in various industries and has the plunging depths from 0 mm to 0.3mm are applied,㻌
rapidly become one of the main processes for joining light when the tool shoulder completely contacts the surface of
materials. Strength of welded joints is a significant issue in upper sheet the rupture load indicates peak values. At this
FSSW researches. Though some reports are published the time, the fracture mode is transited to the thickness
which the key issue is optimization of tool geometry to direction. When further plunging is applied, which is over
obtain the high strength [1-2], the theoretical concept for 0.3 mm the rupture load decreases linearly with the
the tool design has not been established yet. In this study plunging despite the same fracture mode because of the
effects of tool geometry on strength of friction stir spot thinning of the upper sheet. As shown in Figure 2, same
welded joints were investigated in terms of the view of relations are seen using No. 2 and No. 4 tools. However,
material flow during FSSW. the rupture load of welded joints produced by No. 4 tool is
much higher than the other two tools.
2. Experimental Procedure 㻌 㻌 Figure 3 shows the cross sectional macro images of the
The material used in this study was AZ31 with 1.6mm welded joints before and after cross tensile test. When No.
thickness. Figure 1 shows a schematic illustration of the 2 tool is used although width of stir zone is slightly bigger
tool geometries. No.1 is a conventional shape which has a than that No.1 tool is used, the unbonded interface is longer
flat shoulder with 10mm diameter and M4 threaded probe. than No. 1 tool. However, the angle formed by hooking and
No.2 has a concave shoulder with 7.5 degree to produce the interface between the upper sheet and the lower sheet is
inward material flow efficiently beneath the rotating sweep. Thus, the interface would not open when cross
shoulder. No.3 has a bigger shoulder with 15mm diameter tensile test is performed and rupture load is increased.
to increase heat input and be expected to make the stir zone When No. 3 tool is used pronounced hooking is formed and
bigger. No.4 tool has a combined shape of No.2 and No.3 tip of the interface between upper and lower sheet toward
tools to benefit from synergetic effects, and spiral slit is the probe keyhole. Thus, the interface is easily opened
fabricated on the shoulder to produce another material flow when the cross tensile test is carried out and rupture load is
beneath the tool shoulder. All tools are designed on the very weak. In direct contrast, when No. 4 tool is used a㻌
basis of material flow models proposed in previous much bigger stir zone is formed than other tools and
researches to produce the large or optimized bonding area rupture load is very high. Although the interface looks like
[3]. A constant plunging rate and dwell time were applied, a probe keyhole, propagation is in the through thickness
which were 2.5mm/s and 4s, respectively. Tool rotational direction. Spiral slit and concave taper on rotating shoulder
was varied from 1000rpm to 3000 rpm. The shoulder produce a strong inward flow. It is presumed that this
plunging depth below the upper sheet surface was varied material flow facilitates stir zone growth and bonding the
from -0.2 mm to 0.65 mm. Cross tension tests and Tensile interface between upper and lower sheet.
Shear tests were carried out according to the Japanese Figure 4 shows the relation between tensile shear load
Industrial Standard (JIS) Z3137 and JIS Z3136 respectively. and plunging depth when No. 2 and No. 4 tools, which
were confirmed increasing of rupture load in cross tensile
3. Results and Discussion test and No. 1 tool are used. In the case that No. 1 and No.
Figure 2 shows the relation between cross tensile load 4 tools are used, rupture load is increased with increasing
and plunging depth in using No.1, No.2 and No.4 tools. In plunging depth. Fracture mode transits from shear direction,
the case that No.1 tool is used, when the plunging depths which cut across the stir zone to mixed mode of thickness
under 0mm are applied the rupture load increases with the direction and zone direction at around 0.3mm plunging
plunging depth and the rupture pass traverses the stir zone depth. As in the case of cross tensile test, rupture road is
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* Hiroshima University, Hiroshima, Japan Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
** University of Toronto, MSE, Toronto, Canada Japan
28
Effects of tool geometry and process conditions on material flow and strength of friction stir spot welded joints
dramatically higher than other two tools in all plunging and lower sheets would be bonded like friction pressure
depths when No. 4 tool is used. As shown Fig. 4, width of welding. High strength joints are produced stably in every
stir zone is much larger than No. 1 and No. 2 tools. And rotational speed by using No. 4 tool. By comparison with
fracture mode is a mixed mode cutting across stir zone with the higher rotational condition using No. 3 and No. 4 tool,
thickness direction. Thus fracture load is dramatically it is presumed that tool slippage occurs at the contact
increased when No. 4 tool is used. interface between rotating shoulder and material. And
Figure 5 shows the relation between tensile shear load efficient material flow just beneath the rotating shoulder
and tool rotational speed when No. 1, No. 3 and No. 4 tools would not occur when No. 3 tool is used under higher
are used. When No. 1 tool is used fracture load increases rotational speed. However, No. 4 tool has a concave taper
with the tool rotational speed. Fracture mode is in the shear inside of the tool shoulder and spiral slit outside, material
direction at all rotation speeds. This means fracture load beneath tool shoulder is forced to transport inwards. Thus
depends on the width of the stir zone. It is well known that strong inward material flow is produced despite high
the growth of stir zone facilitates with increasing tool rotational condition using big size shoulder.
rotational speed during FSSW using general tool shape.
Thus fracture load increases with the tool rotational speed
when No. 1 tool is used. In direct contrast, fracture load
decreases with increasing tool rotational speed when No. 3
tool is used. When No. 3 tool is used under low rotation
condition a hook is not formed and the interface between
upper and lower sheets are bonded widely. Because
material flow around probe periphery does not occur
enough since heat generation is much lower than using
higher rotational condition. However, big size shoulder
produces efficient material flow beneath the rotating
shoulder. Furthermore axial force during FSSW is higher
when No. 3 tool is used. Thus the interface between upper
Fig. 2 Relationship between plunging depth and Fig. 4 Relationship between plunging depth and tensile
tensile shear load of friction stir spot welded shear load of friction stir spot welded joint made
joint made using different tool geometries. using different tool geometries.
29
Transactions of JWRI, Vol.39 (2010), No. 2
4. Conclusions
Novel tool design which is based on material flow models
during FSSW is proposed in this study. This tool is designed
to facilitate the inward material flow beneath a rotating
shoulder and succeeds in increasing both cross tensile load
and tensile shear load drastically.
References
[1] H. Badarinarayan, et al.: Effect of tool geometry on static
strength of friction stir spot-welded aluminum alloy, Int J Mach
Tools Manuf 49 (2009), pp. 142–148.
[2] Y. Tozaki, et al.: Effect of tool geometry on microstructure and
static strength in friction stir spot welded aluminum alloys, Int J
Mach Tools Manuf 47 (2007), pp. 2230–2236.
[3] S.Horie,et al.: Experimental investigation of Material Flow
Fig. 5 Relationship between rotational speed and tensile during Friction Stir Spot Welding, Science and Technology
shear load of friction stir spot welded joint made
using different tool geometries.
30
Transactions of JWRI, Vol.39 (2010), No. 2
31
Study on the formation mechanism of pore in the intersection of VPPA and FSW
4. Discussion
Fig. 2 Tension specimen Comparing the results of X-ray radiography inspections,
pores can be found in the overlapping seam in the form of
3. Results BW+, but there are no pores in the overlapping seam in the
㻌 Table 2 shows the results of X-ray radiography form of SIP+. The main difference between BW+ sample
inspections. and SIP sample is that there are two initial surfaces of the
BW+ sample. Conclusion that the oxide on the sample
Table 2 Results of X-ray radiography inspections surface leads to the formation of porosity in the intersection
BW+ BW- SIP+ SIP- of the BW+ sample can be drawn. With loosened structure,
the oxide can adsorb water vapor easily. The water vapor
defect Pores No pore No pore No pore
can react with the molten metal, and produce hydrogen in
the arc welding process.
Fracture surfaces of the tensile specimens are shown in Figure 5 is the enlarged picture of a tensile fracture, and
Fig. 3(a). There are a lot of circular shape pores located in it is obvious that some particles are located in some pores.
the neighborhood of fusion line at the root of the welding The particles can be identified as aluminium oxide by EDX.
seam. The pores can be identified as hydrogen pores due to Some fracture surfaces appeared as wave forms, as shown
their smooth inner surface. in Fig. 6, is identical with the fracture which tensile
Figure 3(b)show the enlarged fusion-zone of the tensile specimen bearing zigzag line [6]. Zigzag line is recently
fracture, and no pores but many dimples are observed. hypothesized to be due to entrapped oxide particles seeping
from the specimen surface during welding [7], which can
further confirm that the oxide is the source of hydrogen.
32
Transactions of JWRI, Vol.39 (2010), No. 2
33
Transactions of JWRI, Vol.39 (2010), No. 2
†
Simulations of weld pool dynamics and theis visualization
KEY WORDS: (Arc welding) (Laser welding) (Hybrid welding) (Fluid dynamics) (Animation)
㻌 㻌 㻌 㻌 㻌 㻌 㻌 㻌 㻌 㻌 (Bead formation) (Alloy element mixing)
1. Introduction
Numerical simulations of weld pool dynamics in GMA
welding and laser-GMA hybrid welding can show
information of fluid flow characteristics and alloy element
mixing in detail. The necessary condition for the validity
of simulation results is, however, the accurate description
of complex welding phenomena occurring during heating,
melting, solidification and cooling. This paper introduces
effective ways to describe the heat flux distribution from
arc plasma to the base metal in GTA and GMA welding,
and the laser heat transfer through the surface of keyholes
with arbitrary shapes. The simulation result is visualized
by the animation technique and some meaningful Fig.1㻌 Arc heat flux and bead formation under various
snapshots and compared with the experimental one to shielding gas flow rates in GHTAW
verify the validity of the developed model. The whole
simulation model is then applied to GHTA welding in Figure 2 shows the arc light emissivity, which is the
vacuum, GMA welding of pipes and laser-GMA hybrid result of Abel inversion and a Gaussian fitted line for the
welding of thick plates to understand the weld pool front and side views in GMA pipe welding. To conduct
dynamics of various welding conditions. the Gaussian fitting, the effective radius should be
obtained by taking the integral of arc light emissivity.
2. Arc Welding: GHTAW and Pipe Welding
Previous simulations of arc welding processes
generally adopted the axis-symmetry arc heat flux model
[1]. These studies, however, do not consider the effect of
various conditions such as ambient pressure and joint
shape, which can deform the arc plasma characteristics. In
this study, the arc heat flux model was obtained from the
Abel inversion method [2], and applied for weld pool
㻌㻌
simulations to compare with the results of the generally (a) Front (b) Side
adopted axis-symmetric arc heat flux model in numerical Fig.2㻌 Result of Abel inversion of arc light intensity in
simulations of arc welding processes. 㻌 㻌 㻌 㻌 GMA㻌 pipe welding
Figure 1 shows the arc heat flux distributions
measured by Abel inversion of CCD images of arc light Figure 3 shows the experiment and simulation results
intensity under various shielding gas flow rates in for the different welding positions. Due to the gravity
GHTAW (gas hollow tungsten arc welding), the effect, molten pool behavior affects the variable surface
corresponding characteristic weld pool flows and the profile shapes for different welding positions. In flat and
resulting bead shapes. overhead positions, the molten pool circulated
It is clearly visible that the increase of shielding gas counterclockwise, which makes a uniform bead along the
flow rate results in the intensified and concentrated arc welding direction.
heat flux, which stimulates the strong downward flow
along the center axis. Accordingly the simulated and In the vertical position, however, the molten pool
experimental bead shapes are in a very good agreement cannot circulate well because the gravity force moves the
for different conditions. molten pool to the rear direction, which makes an unstable
bead shape like humping.
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* Department of Mechanical Engineering,KAIST, Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
㻌 Daejeon,Korea Japan
34
Simulations of weld pool dynamics and its visualization
1.8s Acknowledgement
(a) Temperature (b) Time (c) Cr profile Support by POSCO, the Brain Korea 21 project, and a
Mid-career Researcher Program through NRF grant
Fig.4 Simulation results in longitudinal cross sections funded by the MEST (No. 2010-0027749) is gratefully
during hybrid welding acknowledged.
35
Transactions of JWRI, Vol.39 (2010), No. 2
36
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Laser welding simulation) (CIP finite volume method) (Multiphase flow) (One fluid model) 㻌
㻌 㻌 㻌 㻌 㻌 㻌 㻌 㻌 㻌 㻌 (Marangoni convection)
37
Numerical simulation of laser welding processes with CIP finite volume method
(a) t = 0.125 s.
38
Transactions of JWRI, Vol.39 (2010), No. 2
5. Conclusions References
We carried out the laser welding simulation using a one [1] K. Yokoi: J. Comput. Phys. 226 (2007), pp.1985-2002.
fluid model by CIP finite volume method. Although the [2] F. Xiao: J. Comput. Phys. 195 (2004), pp.629-654.
model considered only thermocapillary and surface tension [3] M. S. Gross, I. Black and W. H. Muller: J. Phys. D: Appl.
Phys.36(2003), pp.929-938.
force as a surface force models, the representative behavior
[4] M. Lappa: CMC, vol.2, no.1 (2005), pp.51-64.
of the laser welding was obtained. Therefore, it is [5] M. Von Allmen: Springer-Verlag, Berlin (1987).
concluded that the model can be applied to practical laser [6] J. F. Ready: Academic Press, New York (1971).
welding problems. In order to investigate the details of the [7] S. I. Anisimov, Ya. A. Imas, G. S. Romanov and Yu. V.
flow of the inside the molten pool, the high resolution Khodyko: National Technical Information Services (1971).
simulation is indispensable. So we parallelized the model [8] W. Hirt et al: 4th Int. Conf. Washington DC (1985).
by means of MPI. After some improvement the numerical
model, massively-parallel computers with more than one
hundred CPUs will be used. Thus, in near the future, details
of the laser welding processes will be revealed and it will
contribute to controlling the residual stress on laser welding
processes.
39
Transactions of JWRI, Vol.39 (2010), No. 2
LAURENT D’Alvise*
40
Modelling of the electron beam welding process applied to aircraft engine components
References
[1] P. Ferro, A. Zambon and F. Bonollo, Investigation of elec-
tron-beam welding in wrought Inconel 706 experimental
and numerical analysis, Materials Science and Engineering
A, vol. 392, pp. 94-105, 2005.
[2] A. Lundbäck, Finite element modelling and simulation of
welding of aerospace components, 2003, Licentiate thesis,
Lulea University, Sweden.
[3] R.H. Wagoner and J.-L. Chenot, Metal forming analysis,
Cambridge University Press, UK, 2001
41
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Resistance Spot Welding) (fluid Flow) (Heat Transfer) (Electromagnetic Stirring) (Modeling)
(Magnetofluidodynamics)
2. Numerical Model
A standard RWMA CLASS II electrode (MPE-25Z
CMW® 328) with a flat end surface was utilized in this
work. The material to be welded was 1.6mm thick mild
steel sheet, the welding current was 10,200 ampere, the
welding time was 12 cycles (0.2 s), the ambient temperature
was 21ć, and the temperature of cooling water was 17.8ć.
In order to reduce the complexity of the coupled multi-
physics process, the molten metal in the nugget was
assumed to be incompressible, viscous, laminar, and with
Newtonian fluid. Gravity’s effect on fluid flow was ignored (b)
[1]. The electromagnetic field was viewed as quasi-stable
for low-frequency welding current [2, 3].
The governing equations, boundary conditions,
material models and numerical method are given and
validated in our previous papers [4, 5].
ANSYS/Multiphysics and its parametric design language
were used to realize the coupling of the electric, magnetic,
thermal and flow fields.
42
Visualization of fluid flow and heat transfer in resistance spot weld nugget
Temperature/oC
1200 1200
1000 Without magnetic field 1000
With magnetic field
800 800
600 600
400 400
200 200
Fig. 2 Calculated nugget with traditional electro-thermal 0 0
model. 0.00 0.05 0.10 0.15 0.20 0.25 0.30
Welding time/s
Acknowledgement
This research is supported by National Natural Science
Foundation of China (Grant No. 50705059)
References
[1] J.A. Khan, L.J. Xu, Y.J. Chao, and Kirkland Broach:
Numerical Heat Transfer Part A, 37 (2000), pp. 425-446
[2]Wang Q., Li G.D.: Basic Theory of Electromagnetic, Beijing,
Fig. 3 Calculated nugget with the MHD model. Science Publishing House (1998)
[3] Jia R.G. and Xue Q.Z.: Electromagnetics, Beijing, Higher
Education Publishing House (2003)
[4] Y.B. Li, Z.Q. Lin, S.J. Hu and G.L. ChenO: Journal of
Applied Physics, 101 (2007), pp. 053506
[5] Li, Y.B., Lin, Z.Q., Lai, X.M., Chen, G.L., Zhang, K.: Science
China Technological Sciences, 53 (2010), pp. 1271-1277
43
Transactions of JWRI, Vol.39 (2010), No. 2
Development of welding method for wide gap lap joint of steel sheet
†
using laser welding with hot-wire
KEY WORDS: (Hot-wire) (Laser welding) (Wide gap) (Lap joint) (Steel sheet)
1. Introduction which make the molten pool having a little larger width
In recent years, high strength / ultra-high strength steel than the filler wire diameter, 1.5mm for the spot
sheets are widely used for an automobile body since the diameter and 1.5 m/min for the welding speed in this
demand to reduce automobile body weight and energy study.
consumption is increasing rapidly, moreover lap joints of (2) The adequate wire feeding position and angle which are
high strength / ultra-high strength steel sheets by using laser the anterior region of the molten pool created posterior
welding for body construction is investigated to improve to the penetrated hole, 3mm for the feeding position
the body stiffness. However it is difficult to predict and IURPWKH ODVHU VSRW FHQWHU DQG Û IRU WKH ZLUH IHHGLQJ
control precisely springback behavior during press forming angle in this study.
of high strength / ultra-high strength steel sheets, then a (3) The adequate wire current which makes wire melting in
large gap between lapped sheets is created. In general, laser the region from the upper sheet surface to the gap
welding has only small gap tolerance and the laser welding between the lapped sheets, 102 ~ 116 A in this study.
speed with a filler wire to fill the gap must decrease. In this
study, the new laser welding process, which has a high Table 1 Welding conditions.
welding speed with feeding a filler wire and large gap
tolerance, is proposed and investigated by combining laser
welding and the hot-wire system which has the ability to
control filler wire melting independently from the main heat
input.
2. Experimental procedure
The in-situ observation was carried out using a high-
980 MPa class steel sheets (JIS-G3135-SPFCY, t: 1 speed camera to investigate filler wire melting phenomena
mm) and a 590 MPa class filler wire (JIS-Z3321 YGW23, I and weld bead creation phenomenon during welding. The
1.2 mm) were used. 200 mm long and 50 mm wide observation and evaluation of the bead surface and the
specimens were used for basic investigation, and 220 cross section, and the tensile test were performed after
mm long and 100 mm wide specimens with a 30 mm lapped welding.
region were used for tensile tests. A gap between lapped
sheets was fixed as 1 mm by inserting shim tapes. The weld 3. Result and discussion
bead length was 180 mm and tensile specimens having 30
mm width were cut out in the above welded lapped sheets. Figure 1 shows the high-speed image output during hot-
Table 1 shows the welding conditions. A 3kW fiber wire laser lap welding under the optimum welding
laser, a laser head with a 400 mm focal length and a hot- condition. Figure 2 shows the schematic illustration of
wire heating system produced by Bab-Hitachi Industrial Co. welding phenomenon during developed hot-wire laser lap
were used. The welding speed, the laser power, the filler welding. A penetrated hole initiated and weld metal could
wire feeding speed and the laser irradiation angle were not fill the hole during laser welding without filler wire,
IL[HG DV PPLQ N: PPLQ DQG Û UHVSHFWLYHO\ then the sound bead and joint could not be created. The
The laser spot diameter, the wire feeding angle, the wire developed hot-wire laser welding method can make stable
feeding position and the wire current were changed as wire feeding to fill the hole created by laser irradiation, and
experimental parameters. Filler wire was fed from then the sound joint with a 1 mm gap could be created by
backwards into a welding region. the proposed welding method. The following conditions are
necessary to make stable filler wire feeding and the sound
(1) The adequate laser spot diameter and the welding speed joint.
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* Hiroshima University, Hiroshima, Japan Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
** Mazda Motor Corporation, Hiroshima, Japan Japan
44
Development of welding method for wide gap lap joint of steel sheet using laser welding with hot-wire
4. Conclusions
The hot-wire laser welding method was proposed for the
Fig.3 Cross-sections when wire current is varied. lap joint of the high tensile / ultra-high tensile strength
sheets with a wide gap, and its welding phenomenon was
Figure 4 shows the effect of the wire feeding angle on investigated precisely by using a high-speed camera, then
the optimum wire current. When the wire feeding angle the optimum welding conditions were obtained in this study.
EHFDPH WRR ORZ Û DQG Û LQ WKLV VWXG\ ILOOHU ZLUH The obtained conclusions are as follows.
touched the upper sheet surface at the back of the molten (1) The hot-wire laser system makes the penetrated hole
pool, then stable wire feeding could not be performed. and the molten pool by laser irradiation firstly, and then
When the wire IHHGLQJ DQJOH ZDV WRR KLJK Û LQ WKLV filler wire heated up by energization fills the penetrated
study), the filler wire tip melted down frequently since the hole in the proposed hot-wire laser lap welding method.
reflected laser beam irradiated unsteadily on its surface. The optimum conditions for the 1 mm gap are 1.5mm of
:KHQWKHZLUHIHHGLQJDQJOHZDVÛLQWKLVVWXG\WKHILOOHU the spot diameter, 3 mm of the wire feeding position,
wire tip touched stably the molten pool posterior to the Û RI WKH ZLUH IHHGLQJ DQJOH DQG a $ RI WKH
penetrated hole, then the optimum wire current region wire current for 3kW of laser power, 1.5 m/min of the
became wider. welding speed, 3.8 m/min of the wire feeding speed, 1
45
Transactions of JWRI, Vol.39 (2010), No. 2
mm of the base sheet thickness and 1.2 mm of the filler upper sheet surface to the gap between the lapped sheets
wire diameter. are necessary for the stable welding phenomenon and
(2) The laser spot diameter and the welding speed which sound bead creation.
make the molten pool having a little larger width than (3) The stable and high tensile shear strength (500 ~ 600
the filler wire diameter, the wire feeding position and MPa) can be obtained over the wide wire current region
angle which is the anterior region of the molten pool (102 ~ 120 A) using the optimum welding conditions
created posterior to the penetrated hole, and the wire based on the above investigation.
current which makes wire melting in the region from the
46
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Hot wire) (Fillet welding) (Laser welding) (High efficiency)
2. Experimental procedure
Mild steel plate (SS400, thickness: 9 mm) and filler
wire (JIA Z3321 YGW, I 1.4 mm) were used. Table 1
shows the chemical compositions of used materials.
Table 2 indicates the welding conditions. The
conditions relating with laser and filler wire were changed
as experimental parameters. A fiber laser was used as a heat The high speed camera images for each of the wire
source. The laser power and the welding speed were fixed. feeding distances forming the laser spot are indicated in Fig.
Spot diameters of laser were controlled by the distance 2. The stable melting can be confirmed if the fire is fed into
from the focal length of a laser head. A laser head and a 2 mm of the position from the spot center. However, the
wire torch were leaned Û WR WKH vertical direction of the filler wire is melted before touching the base material due
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* Hiroshima University, Hiroshima, Japan Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
** IHI Inspection & Instrument, Kanagawa, Japan Japan
*** IHI Corporation, Kanagawa, Japan
47
Development of a high-efficiency / high-quality hot-wire laser fillet welding process
Fig. 1 High speed images on each wire current. Fig. 4 Cross sections of weld bead on each spot diameter.
Spot diameter : 7.5mm, Wire feeding angle : 55 °
to the direct irradiation of laser, when the wire is introduced Wire feeding position: 1.5mm, Wire current :
at the center of the laser spot. On the other hand, the wire is 0~149A
not able to be inserted into the molten pool for more than 3
mm of wire feeding position. These result means that the
filler wire should be introduced into the molten pool
backward without the laser irradiation
In addition, we also investigated the wire melting
phenomena as the function of the wire feeding angle. In the
FDVH RI DQG Û RI WKH IHHGLQJ DQJOH WKH FRQWLQXRXV Fig. 5 Cross sections of weld bead
feeding and the stable melting of filler wire were obtained. on various wire feeding speeds.
+RZHYHU Û RI WKH IHHGLQJ DQJOH led to form the poor
geometry of the bead since the wire tends to touch the solid Regardless of the filler wire addition, the length increases
part at the backward part of the molten pool due to the wire with increasing spot diameter. Besides, those values of no
softening by heating. wire samples are smaller than those of the wire added ones.
As the second step, the phenomenon of bead formation It can be assumed that this is caused by the reflection of the
was evaluated. Figure 3 shows the change of average leg laser beam. The laser beam has a great tendency to reflect
length with spot diameter of laser. Average leg length was on the molten pool surface due to high optical reflectivity of
obtained from both the lengths of vertical (L 1 ) and liquid metal. Thus, the reflected beam contributes to
horizontal (L 2 ). The broken line indicates the laser melting of the base material since the beam irradiates the
irradiated length calculated from the spot diameter. surroundings of the distended geometry of the molten pool.
Figure 4 indicates cross sections of the beads welded
Spot diameter: 7.5 mm, Wire current: 110 A, under identical condition shown in Fig. 3. In the case of 3
Wire feeding speed: 1.8 m/min, Wire feeding angle: 55° mm of the spot diameter, the penetration is deeper
compared with another samples because of the high energy
density. If we apply the laser spot diameter which is more
than 4.5 mm, the dilutions of the base materials are quite
low. In addition, defects like overlaps and undercut at the
welding toe never form in any spot diameters. However,
lack of fusion at the root observed in the samples fabricated
under 6.0 and 7.5 mm diameter of the laser spot due to the
a) 0 mm b) 2 mm
low heat input corresponding to large spot diameter.
Fig. 2 High speed images on each wire feeding position.
Figure 5 also shows cross sections as the function of
wire feeding speed. There is not much difference between
the leg lengths of the weld bead. However, the cross section
geometry transits from triangle to sector. Furthermore,
increasing the wire feeding speed induces to the formation
of larger lack of fusion at the root. Thus, there is a limit to
improving the leg length by the laser beam reflection
changing with only the wire feeding speed. We should
consider the factor interaction between laser energy density,
wire feeding speed and welding speed.
The welded joint fabricated under the optimized
conditions has following properties; average grain size in
the coarse grain zone: 60 μm, dilution ratio: a few percent,
thickness of heat affected zone: 1 mm.
These results reveal that this process includes various
factors influencing on the welding phenomenon. However,
Fig. 3 Leg length change with spot diameter of laser. the parameter optimization contributes to the stable melting
of base materials and the filler wire and the obtained welds
48
Transactions of JWRI, Vol.39 (2010), No. 2
indicate attractive properties, namely low heat input and (2) The filler wire should be fed into the molten pool
low dilution. backward without the laser irradiation.
(3) The leg length increases by the addition of filler wire
4. Conclusions compared with the non filler welding due to the laser
In order to develop the high efficiency / high quality beam reflection at the molten pool surface.
fillet welding technique, the welding phenomena on the hot- (4) The formation of lack of fusion at the root must be
wire laser fillet welding were investigated. Obtained caused by the excessive wire feeding and the
conclusions are as follows. insufficiency of the heat input by laser.
(1) The melting of filler wire depends on wire current and (5) The parameter optimization leads to the stationary
there is an optimum vale of the current at each wire welding phenomenon. In addition, the obtained weld
feeding speed. has attractive properties, namely low heat input and low
dilution.
49
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Metal) (Plastic) (Ceramic) (Laser direct joining) (TEM image) (LD laser)
1. Introduction
Many materials including metals, plastics and ceramics
have been widely used in many industrial applications such
as automobiles, aircrafts and electronic devices. Joining
of these dissimilar materials is necessary and important
from a manufacturing viewpoint. The features of metals
include high strength, high toughness, high heat
conductivity and high heat resistance. Plastics are
characterized by lightweight, high corrosion resistance and
excellent formability. Polyethylene terephthalate (PET) is
known as a typical engineering plastic and commercially
available bottles. On the other hand, ceramics are used
owing to their high hardness and/or good heat resistance.
Silicon nitride ceramics (Si 3 N 4 ), a chemical compound of
silicon and nitrogen, is utilized as a bearing for engines or a
Fig 1 LAMP joints before and after tensile shear test.
rotor blade of turbine.
The dissimilar joining is generally performed using shown in Fig. 2. The TEM image demonstrates that the
adhesive bonds (glues) or mechanical fastering such as metal and the plastic are tightly bonded on the atomic or
bolts and rivets. However, these joining processes have molecular sized level. The base metal and the
some problems in terms of environmental restriction of intermediate layer were identified to be fcc gamma (J)
volatile organic compound (VOC) emission and mass phase and 5 nm-thick Cr-oxide film from these diffraction
production. Therefore, the authors have developed a new patterns, respectively. It was therefore revealed that a
laser direct joining method between metals and plastics strong joint could be produced by atomic, nanostructural or
which is named Laser Assisted Metal and Plastic joining molecular bonding of the metal and the plastic through the
method (LAMP joining method) [1]. oxide film on the metal plate surface. Not only the anchor
The LAMP joints between engineering plastic PET and (mechanical bonding) effect but also Van der Waals
stainless steel SUS304 before and after the tensile shear test interaction force and chemical bonding were considered as
are shown in Fig. 1. The top surface of PET was not bonding mechanisms of LAMP joining.
damaged to be as smooth as what it had been before the In this research, LAMP joining method with a 3-kW
laser irradiation. However, many bubbles of sub- diode laser (LD) was applied to join of Si 3 N 4 ceramic and
millimeter sizes were formed inside the plastic near the PET engineering plastic. The LAMP joining between the
joint. The shear load of the joint was about 3000 N, and ceramic and the plastic was carried out to optimize welding
the base PET was elongated sufficiently as shown in the speed. The obtained joints were evaluated by the tensile
lower photo of Fig. 1. Such strong joints were produced shear test. Furthermore, the joints were observed in
by the laser direct joining process with many small bubbles. details through TEM in order to investigate the joining
The LAMP joints were examined with transmission structure.
electron microscopes (TEM), and these examples are
50
Innovation of laser direct joining between metal and plastic
51
Transactions of JWRI, Vol.39 (2010), No. 2
4. Conclusions References
[1] Y. Niwa, Y. Kawahito, S. Kubota and S. Katayama, Proc. of
We have successfully produced new LAMP (Laser-
ICALEO 2007, LIA, Orlando, (2007) (CD).
Assisted Ceramic and Plastic) joints between the
[2] S. Katayama and Y. Kawahito, Scripta Materialia, 59-12
commercially available ceramic Si 3 N 4 and the engineering
(2008), pp.1247- 1250.
plastic PET. The joint possessed over 3100 N shear load
strength in 30-mm-width plastic samples of 2 mm in
thickness. The base plastic was elongated because of the
formation of strong joints. Moreover, the TEM image of
the LACP joint demonstrates that the ceramic and the
plastic are tightly bonded on the atomic or molecular size
level.
52
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (CFRP) (Laser cutting) (Ultra-high cutting speed) (Heat Affected Zone) (Cutting surface)
㻌 㻌 㻌 㻌 㻌 㻌 㻌 㻌 㻌 㻌 (High brightness laser)
1. Introduction
Carbon fiber reinforced plastic (CFRP) with high Table 1 Material properties of CFRP used in experiments
specific strength to weight ratio, outstanding fatigue
resistance and no corrosion as one of composite materials
has been widely applied in automobile, aerospace and
electronics, where it is necessary to reduce energy
consumption and CO2 generation. However, CFRP is
difficult to cut because of the high hardness of carbon fiber
that made PITCH carbonized at high temperatures. The
main cutting techniques of CFRP are abrasive water-jet
machining and mechanical cutting operations by contact
process between tool and material such as micro-cutter
machining, milling machining. But they have several
disadvantages such as moisture absorption, high tool wear
because of mechanical loads and thermal loads, high costs Fig. 1 Schematic experimental set-up for cutting of CFRP
and limited cutting speed. In recent years, laser cutting
experiments with high power density heat source as one of
non-contact process have been conducted to investigate
cutting characteristics of CFRP using pulsed Nd:YAG laser,
disk laser and CO2 laser[1, 2]. During the laser cutting of
CFRP with high laser beam absorptivity, however, a㻌 wider
heat affected zone could be formed because of an increase
in laser heat input at a range of lower speed and also was
affected by material properties of CFRP and laser cutting
parameters including the laser power, the cutting speed, the
pressure of assist gas and the focal position. In this study, Fig. 1 Schematic experimental set-up for cutting of CFRP
therefore, with the objective of obtaining the better cut
quality and narrower HAZ width, laser cutting experiments beam of 8 mm䂩mrad (BPP) was delivered by the optical
were conducted to establish the proper laser cutting fiber of 0.2 mm core diameter. The laser scanner head was
conditions for various CFRP sheets at ultra-high cutting employed in order to accomplish ultra-high cutting speed,
speed using high brightness CW disk laser. and the laser beam was focused by the focal lens of 291.5
mm. The spot size of the laser beam was about 0.3 mm at
2. Experimental procedures the focal point.
Table 1 shows material properties of CFRP sheets used Table 2 indicates laser cutting conditions used mainly
in the experiments. Each CFRP was made by considering for cutting of CFRP. Laser cutting parameters were the
different material properties which include plastic type, laser power of 1 kW, the cutting speed of 500 – 5000 mm/s,
volume of fiber (%), fiber type, manufacturing process㻌 and laser irradiated number of 1 – 80 passes, time interval after
thickness of CFRP. A schematic experimental set-up for laser irradiated of 0 or 1 second, the focal position of ±0
laser cutting of CFRP is given Fig. 1. The CW disk laser mm. Cutting phenomena and behavior of the laser induced
with a maximum power of 16 kW was utilized and the laser plume during the laser cutting were observed by using two
53
High brightness laser cutting of CFRP
high speed video cameras at 10,000 F/s and 5,000 F/s for
laser irradiated part and laser induced plume, respectively.
A He- Ne laser with the maximum power of 50 mW and the
54
Transactions of JWRI, Vol.39 (2010), No. 2
55
Transactions of JWRI, Vol.39 (2010), No. 2
56
Characteristics and mechanism on the distortion of friction stir welded aluminum alloy sheet
input from the shoulder was 75% of the total, while the model 1 was consistent with the FSW experimental results
other was generated by the pin. The heat supplied by the in shape, while the distortion shape from the model 2 was
tool shoulder was considered as a surface heat flux which is absolutely opposite. The offset between the highest and the
axi-symmetric and increases linearly with radius, and the lowest position of sheet was 7.41mm and 8.63mm in the
heat generated by the pin was assumed to be a uniform model 1 and the model 2 respectively. These results
volume heat flux [6, 8-9, 17-18] . indicated that the introduction of stir tool’s mechanical
The heat transfer condition on the sheet surfaces which loads had the effect of distortion reduction. The distortion
were not in contact with the fixtures was set as the curves on the line of X=40 mm, X=320 mm, X=560 mm,
convection boundary condition, while on the contacting Y=-310 mm, Y=-116 mm and Y=0 mm on the top surface
surfaces the heat transfer was treated in a way of contact of the sheet from the computation with the model 1 and
heat conduction. The contact conductivity was assumed to experimental results were listed in Fig. 5. The results
be simply temperature-dependent [19]. showed that the distortion trends of the sheet by simulation
The restriction effect of the welding fixtures on the and experimental results were consistent, and this
sheet was set as a contact model, where the condition on consistency was compatible over the whole plate, not for
contacting surfaces was described as Coulomb friction with some local positions. The largest error between the
a coefficient of 0.3, and the interaction force in the normal deformation values was less than 20%, which was
direction was set under the principle of “pressure-over- acceptable in welding mechanical analysis. The distortion
closure”[20]. The down force of the stir tool was imposed comparison confirmed that the model 1 was a reasonable
as a uniform surface pressure onto the upper surface of the model for distortion analysis of FSW.
sheet where the stir tool covered. The working torque was
simplified as shearing friction between the shoulder
surfaces and welded materials, and was treated as
circumferential body forces, and applied on the uppermost
layer integration points of the panels which lay in the region
covered by tool shoulder [21]. Model 2 only has thermal
load during mechanical analysis.
At the end of mechanical simulation, the restrictions of
welding fixtures were removed. Thus the residual distortion
of sheet appeared under unclamped condition.
57
Transactions of JWRI, Vol.39 (2010), No. 2
58
Transactions of JWRI, Vol.39 (2010), No. 2
2. Measurement
59
The influence of the solid state phase transformation on welding deformation of low alloy high strength steel
3. Calculation
We did a sequential coupled analysis on the same
model based on ABAQUS, taking account of the factor of
phase transformation. During the welding, diffusion and no Fig. 5 Comparison of calculation and experiment
diffusion phase transformation happens in the region near
the weld at different cooling speeds[2]. The diffusion phase Figure 5 shows the calculated result compared with
transformation can be described by the following the experiment result. Figure 6 shows the calculated
Kamamoto model[3]: displacement in the mid point compared with the
Ts T n experiment, with another plot not taking account of phase
X 1 exp[k ( ) ] (1) transformation. We can see the trend of deformation
Ts T f conforms to the experiment, but the values differ a lot. The
X represents the volume percentage of the new phase. displacement taking account of phase transformation is
T s and T f represent the moment phase transformation starts smaller than that of not taking account of phase
and finishes, which are determined by the cooling speed transformation.
and continuous cooling transformation chart as shown in
Fig. 4. The parameters k and n come from previous
experiment result.
4. Conclusions
The conclusions of this study are summarized as follows.
(1) In one-side fixed welding of thin plate, the unfixed side
warps a lot, especially at the position in the middie.
Fig. 4 The CCT chart of AH3201 (2) The deformation appears to be an increase-decrease-
increase process, either in the experiment or in the
60
Transactions of JWRI, Vol.39 (2010), No. 2
calculation, no matter if we take account of phase [2] Zhao Hongzhuang, Xu Hong, Liu Xianghua and Wang
transformation. Guodong: Review of Numerical Simulation on the Process of
(3) Phase transformation from austenite to phases at a lower Heat Treatment. Heat Treatment, 2004,19(1).
temperature decrease the welding transformation. [3] Wang Qingpeng, Li Wushen, Chen Cuixin and Feng Lingzhi:
Analysis on the Model of Phase Transformation in Welding
Acknowledgement Heat Affected Zone. Welding, 2004(11).
This project is support by the National Natural Science [4] Wang Wenxian, Huo Lixing, Zhang Yufeng and Wang
Foundation of China (Grant No. 50975176) Dongpo: Analysis on the Regulation and Mechanism of the
Influence from Phase Transformation Temperature over
References Welding Residual Stress. China Mechanical Engineering,
[1] Xu Jijin, Chen Ligong, Zhang Min, He Defu and Wu Youjing: 2003,14(3).
Experiment and Measuring System of Double Wires
Submerged Arc Welding with A105-Steel Heavy Plates.
Symposium of 11th National Welding Conference, 2005.
61
Transactions of JWRI, Vol.39 (2010), No. 2
62
Optimising residual stress measurements through the use of measurement simulation
63
Transactions of JWRI, Vol.39 (2010), No. 2
References
[1] C. E. Truman and M. C. Smith: Int. J. Pres. Ves. Pip., 86(1)
(2009), pp 1-2
[2] O. Muránsky et al: ASME Pres. Ves. Pip. Conf. (2010),
PVP2010-25
64
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Welding residual stress) (Austenitic stainless steel) (Butt-welded joint) (Inherent strain
method) (Large-bore pipe) (Three-dimensional stress distribution)
1. Introduction
The inherent strain method [1] is a destructive § ] ·§ [ ·
H *kf ] ,[
A1k ¨¨1 ¸¸¨¨1 ¸¸
technique to measure the residual stress accurately by © Z k ¹© R k ¹ (4)
cutting up the welding structure to be measured. The N N
§ S] · § S[ ·
residual stress is represented by Eq.(1) using inherent ¦¦ Aijk cos¨¨ 2i 1 ¸¸cos¨¨ 2 j 1 ¸¸
strain: i 1 j 1 © 2Z k ¹ © 2 Rk ¹
m
(2) V, respectively. Tungsten inert gas (TIG) welding was done
without fixing the ends from 0° to 180°.
The inherent strain method using functional form [2], in 3. Results and Discussion
which the inherent strain distribution was represented by Figures 1 and 2 show the residual stress distributions
the function, enables accurate measurement in spite of for circumferential and axial stress on the inner and outer
fewer measurement points because the distribution of sides in the 0° - 180° cross section. A different bore
inherent strain ^H
` is represented by the appropriate changes the distributions of compressive and tensile stress.
function. This is why the residual stress distribution is more
In this study, the inherent strain method using functional complicated due to the change in magnitude of the bending
form was applied to welded pipe joints. The measurement deformation because the magnitude of thermal expansion
strain was measured at 90° from the welding start position and the shrinkage are different in the circumferential
and then, the residual stress distribution was calculated direction as the bore and wall thickness are each made
assuming that the strain distribution spread uniformly in the larger.
circumferential direction. Here, the third order polynomial Figures 3 and 4 show the residual stress distribution of
function was assumed as an appropriate function for the 300A welded pipe joint. The distributions calculated by
inherent strain distribution. However, the appropriate using two functions with N=3 in Eq.(3) and Eq.(4) are very
function of inherent strain distribution has not been clarified reliable based on statistical and mechanical standpoints [3].
for a large-bore and heavy-walled welded pipe joint. However, more investigation is needed to determine the
Therefore, six kinds of functions were examined to most reliable function. Additional functions with various N
represent the distribution appropriately. values are now being examined.
N N
§ ] ·
i
§ [ ·
j
4. Conclusions
H * ] ,[
f
k ¦¦ Aijk ¨¨1 ¸¸ ¨¨1 ¸¸ (3) The results in this study were summarized as follows:
i 1 j 1 © Zk ¹ © Rk ¹
65
Residual stress measurement of large-bore stainless steel pipes with butt-welded joints by inherent strain method
1.5E+02 2.0E+02
2.0E+02
1.5E+02
1.0E+02
1.0E+02 1.0E+02
5.0E+01
5.0E+01
0.0E+00 0.0E+00
0.0E+00
-5.0E+01 -5.0E+01 -1.0E+02
-1.0E+02 -1.0E+02
-2.0E+02
-1.5E+02 -1.5E+02
-2.0E+02 -2.0E+02 -3.0E+02
1.5E+02 1.5E+02
1.5E+02
1.0E+02 1.0E+02
1.0E+02
5.0E+0.1
5.0E+01
5.0E+01 0.0E+00
0.0E+00
0.0E+00 -5.0E+01
-5.0E+01
-1.0E+02
-1.0E+02 -5.0E+01
-1.5E+02
-1.5E+02 -1.0E+02
-2.0E+02
-2.0E+02 -1.5E+02 -2.5E+02
Fig.4 Axial residual stress of 300A welded pipe joint (90° cross section).
66
Transactions of JWRI, Vol.39 (2010), No. 2
(1) The inherent strain method using functional form was References
applied to large-bore and heavy-walled stainless steel [1] Y. Ueda, K. Fukuda, K. Nakacho and S. Endo: J. Soc. Nav.
pipe with butt-welded joint. Archit. Jpn., (1975), pp.499-507.
(2) The shape of the stress distribution had a more complex [2] Y. Ueda and N.X. MA: Quarterly Journal of JWS, 11 (1993),
change when the bore and thickness were larger. pp.189-195.
(3) Six kinds of functions were examined to provide greatly
[3]K. Nakacho, N. Ogawa and T. Ohta: Quarterly Journal of JWS,
reliable distribution of residual stress. 27 (2009), pp.297-306.
67
Transactions of JWRI, Vol.39 (2010), No. 2
†
Residual stress generated by LBW on HT780
KEY WORDS: (Residual stress) (Laser beam welding) (High strength steel) (Phase transformation)
(Thermal elastic-plastic analysis)
20
40
It is unknown how residual stress is generated in
welding of the high strength steel by laser beam.
In this paper, in order to elucidate how residual stress x
9mm
is generated by LBW on the high strength steel (HT780), an
experiment is carried out and is simulated by a thermal y
300
2. Experiment
20
68
Residual stress generated by LBW on HT780
500
0
z x=30 (mm)
Welding Obverse
direction y
x -500 f
Weld metal -50 -25 0 25 50
y (mm)
(b) Stress component in transverse direction; ıy
Symmetric Fig. 4 Residual stress distribution
condition
HAZ In the case of without PT modelı x is over the yield
stress. However, it is confirmed that the equivalent stress
DJUHHVZLWKWKH\LHOGVWUHVV 2QWKHRWKHUKDQGı x varies
from the yield stress to zero in the case of with PT model.
Fig. 3 Model for thermal elastic-plastic analysis. The experimental result is within the result obtained by the
analysis of with PT model. Considering the accuracy of
Figure 3 shows the analysis model. the stress relaxation method, the experimental result is
The isoparametric solid elements with 8 nodes are substantially simulated by the analysis of with PT model.
used. The stress compRQHQW LQ WKH WUDQVYHUVH GLUHFWLRQ ı y
The half model is adopted. The heat input (the weld (Fig. 4(b)) is noted.
metal) elements are decided by referring to the macrograph In the case of without PT modelı y is tension and the
(Fig. 1). value is around 300MPa. It does not agree with the
The model with considering the phase transformation experimental result which is small compression. On the
by using the above idealized mechanical properties is contrary, the result obtained by the analysis of with PT
defined as ‘with PT’ model. model accurately agrees with the experimental result.
On the other hand, the model without considering the From these results, the validity of the treatment in
phase transformation, i.e., the same mechanical properties mechanics of with PT model idealizing the mechanical
in the heating process are also used in the cooling process, properties with considering the transformation expansion
is defined as ‘without PT’ model. and transformation superplasticity is verified.
The analysis results of these two models are compared The residual stress generated in the weld metal is
to investigate the effect of the phase transformation on the largely affected by the phase transformation in the cooling
residual stress. process although the bead width of LBW (around 5mm) is
extremely narrow. The tensile residual stress generated in
5. Results of Experiment and Analysis the weld metal is largely released by the phase
Figure 4 shows the residual stress distribution transformation.
obtained by the experiment and the analysis.
The stress at the center of the welding direction 6. Conclusions
(x=30(mm)) on the obverse surface is shown in the figure. In order to elucidate how residual stress was generated
The stress component in the welding directioQı x (Fig. by LBW on the high strength steel, one-pass bead-on-plate
4(a)) is noted.
69
Transactions of JWRI, Vol.39 (2010), No. 2
welding of HT780 by laser beam was simulated by the compared with that of existing arc welding.
thermal elastic-plastic analysis with using the idealized (3) The tensile residual stress generated in the weld metal
mechanical properties considering the phase transformation. was largely released by the phase transformation.
(1) The residual stress obtained by the experiment could
be simulated by the analysis with high accuracy. The References
validity of the treatment of the mechanical properties in [1] INOSE K., LEE J-Y., NAKANISHI Y. and KIM Y-C. :
the phase transformation range was verified. QJJWS, 26-1(2008), pp. 61-66.
(2) It was known that the residual stress generated in the [2] SATOH K. and TERASAKI T. : Journal of the JWS, 45-
weld metal was largely affected by the phase 7(1976), pp.46-52.
[3] UEDA Y., KIM Y-C., CHEN C. and TANG Y-M. : QJJWS,
transformation in the cooling process although the bead 2-1(1984), pp. 89-97.
width of LBW (around 5mm) was extremely narrow
70
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Air conditioning compressor) (Welding deformation) (Blade groove width) (Three-point
welding) (Phase transformation) (Thermal elastic-plastic finite element method)
1. Introduction
The FEM simulation of welding residual stresses and
distortions can provide valuable instructions for the
optimization of welding technology and structures. Under
the welding condition, the phase transformation in solid (a) Geometric model
state not only changes the strain-stress, but also affects the 5x108
thermal elastic-plastic
life evaluation for the structure. In order to accurately 4x108 thermo-mechanic-metallurgical
predict the distortion and stress during welding process, the 3x10 8
1x108
caused by phase transformation are also considered [1,2]. 0
71
Structure optimization of air conditioner compressor based on reverse welding deformation model
0.012
Width Change of the groove/mm
0.010
0.008
0.006
(c) Case C (×30)
height 60mm not contacted
0.004
height 220mm contacted Fig. 4 Contours of radial deformation under different
0.002 height 220mm not contacted conditions
height 220mm not contacted with different materials
0.000
0 5 10 15 20 4. Conclusions
Distance from the end/mm According to the simulation results, the following
Fig. 3 The width change of the blade groove with conclusions can be obtained:
different parameters (1) The height of the compressor shell has little effect
on the width change of the blade groove. The width change
The deformation of the shell with different parameters becomes smaller when cast steel is used for the cylinder.
is shown in Fig. 4. Three different cases are discussed: (2) When the shell and the cylinder are in contacted,
Case A, the shell and the cylinder are in contact, and the deformation of the shell is much smaller. And in the
using the same material properties. experiment, the measured cylinder roundness is 377.41ȝm,
Case B, the shell and the cylinder are not in contact, which is in very good with the case C.
using the same material properties.
Case C, the shell and the cylinder are not in contact, the Acknowledgement
cylinder using different material properties. This project is support by the National Natural Science
The results show that, the deformation of the Foundation of China (Grant No. 50975176)
compressor shell seems like a “flower”. When the shell and
the cylinder are in contact, the deformation of the shell is References
much smaller. The cylinder roundness of these three cases [1] Jun-yan LIU, Hao LU, Jun-mei CHEN, Jinyang LIU. Finite
are 229ȝm䚸416.2ȝm and 397.8ȝm. And in the experiment, element simulation of Satoh test taking into account
the measured cylinder roundness is 377.41ȝm, which is in transformation induced plasticity. Transactions of the China
very good with the case C. Welding Institution, 2008, 29(3):105-109.
[2] Jun-yan LIU 䠈㻴㼍㼛㻌 㻸㼁䠈 㻶㼡㼚 -mei CHEN. Experimental
analysis of transformation induced plasticity in 9Cr1Mo steel.
Front. Mater. Sci. China 2009, 3(1):61-66.
[3] Caner Simsir, Cemil Hakan Gür. 3D FEM simulation of steel
quenching and investigation of the effect of asymmetric
geometry on residual stress distribution. Journal of materials
processing technology, 2008, 2007:211-221.
[4] Caner Simsir, Cemil Hakan Gür. A FEM based framework for
simulation Application to steel quenching.
(a) Case A (×40)
Computational Materials Science, 2008, 44:588-600.
[5] Moumni, Z., et al. Theoretical and numerical modeling of the
thermo mechanical and metallurgical behavior of steel. Int. J.
Plasticity (2010), doi:10.1016/j.ijplas.2010.07.002.
72
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (T-E-P FEM) (Welding deformation) (Large-scale structure) (3D shell element) (Parallel
calculation)
73
Study on prediction of welding deformation for large-scale structure by T-E-P FEM using 3D shell element
3. Calculation example
As an example, a typical of the hull structure, called
large-scale hull double bottom structure, is modeling, and
3D shell element T-E-P FEM was used to simulate welding
deformation.
74
Transactions of JWRI, Vol.39 (2010), No. 2
4. Conclusions
The conclusions of this study are summarized as
follows:
(1) Regarding bead-on-plate welding, the calculation
results obtained by 3D solid and shell element T-E-P
FEM are in good agreement.
(2) If mesh size is less than 100mm, calculation results
will not be affected greatly by mesh size
(3) By simulation of a typical of large-scale hull double
bottom structure, the feasibility of 3D shell element T-
E-P FEM was verified.
References
[1] Yu LUO, Hidekazu MURAKAWA and Yukio UEDA: Trans.
Fig. 6 Welding Deformation
Of JWRI, Vol. 26 (1997), No.2.467-475.
75
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Large thick-wall) (Fast prediction) (Stress evolution) (Roundness change) (Simplified models)
Heat input
0.12
-3/[W/(mgć
)]
0.06
¬×10
Ch10-4/[ J/(kg·ć)]
50
0.00 E/GPa
36
0 500 1000 1500 2000
Unit˖mm 2550 Temperature t /ć
2.25 Time t / s Į: Linear expansion coefficient; ı s : Yield strength; E: Elastic modulus; C:
Specific heat; O: Thermal conductivity; ȡ(Density)=7.89×106kg/m3
Fig.1 Structure and groove of welded
Fig.2 Heat input and their duration Fig.3 Material property
cylinder with angle-inserting elbow
3. Comparison of simplified model The regions where the equivalent residual stress under 3
simplified models is up to 80%, 60% and 50% respectively
Three simplified models are established. The welding are compared as shown in Fig. 8. It can be seen that the
order and weld size after the beads is simplified to be 5 and methods of bead and moving heat source simplified have
are described as shown in Fig. 5. The simplification little effect on the overall stress of joint.
methods are given in Table 2. In order to explain the effect of simplification method
The stresses of point A at different locations along the on the stress evolution of joint, the process stress under 3
depth direction of the weld are compared as shown in Fig. 6. simplified models of point A at different moments are
It can be known that, the stress distribution under model A compared as shown in Fig. 9.
is most close to the typical description of stress distribution
of thick-wall vessel in other reference [3]. 4. Fast prediction
Stresses of region B at different locations along the Under the Model C, the computing time under this
vertical direction of weld are compared as shown in Fig. 7. simplified model can be shorten significantly. Therefore,
It can be known that the method of simplified bead is Model C whose bead and moving heat sources are both
obviously better than that of simplified moving heat source simplified is selected for fast prediction of joint stress in
when studying the stress within the same bead plane. this article.
76
Fast finite element stress and deformation prediction for large thick-wall welded cylinder with angle-inserting elbow
125
50mm Model A
5
100 Model B
Depth h / mm
Model C
4 75
125mm
3 50
2 25
1
0
10mm 135 150 165 180 195 210
Stress ³m MPa
Model B
200 Model C
150
100
Stress ³/ MPa
Stress ³MPa
Stress ³MPa
Stress ³/ MPa
100 0
50 -100
50 -20
0 5000 10000 15000 0 150 300 450 600 0 150 300 450 600
0 5000 10000 15000
Distance d / mm Distance d / mm
Time t / s Time t / s
˄a˅Internal surface ˄b˅External surface ˄a˅Longitudinal residual stress ˄b˅Transverse residual stress
Fig.9 Stress cycle on the internal and external weld at point A Fig.10 Stress of the weld with distance from the centerline
For the analysis of deformation characteristics of joints, Figure 12 shows the relative displacement of all points
the roundness change of the left and right ends on the on the elbow outer contour at different moments. At the
cylinder, the roundness change of weld curve, and the beginning of welding, elbow is far from the cylinder, finally
displacement of elbow outer contour are analyzed, the the distance between elbow and cylinder is smaller and the
research area is shown as in Fig. 11. axis curvature of elbow is larger.
Y direction
Elbow outer
contour
Original state
Welding initial
Welding completed
After cooling
Right end
Left end
X direction
Fig.11 Research region of displacement Fig.12 Displacement schematic diagram of elbow axis
Figure 13 shows the roundness change at the two ends Figure 14 expresses the displacement of all the points
of left and right on the cylinder. During cooling, the on the weld at different moments. It can be seen that the
roundness change of the cylinder is just opposite to that in angle between the symmetrical axis of weld and the axis of
the stage of welding. The elliptic deformation of circle at cylinder has an important influence on the deformation of
the left of the cylinder is smaller than that of the circle in the weld.
the right.
77
Transactions of JWRI, Vol.39 (2010), No. 2
0 0 0
330 30 330 30 330 30
Original state Original state
Welding initial Original state
300 60 Welding initial 300 60 300 60 Welding completed
Welding completed Welding completed
After cooling
After cooling After cooling Measured results
270 90 270 90 270 90
78
Transactions of JWRI, Vol.39 (2010), No. 2
1. Introduction
It has been more than three decades since the finite
element method was first applied to the welding simulation.
However, due to the poor performance of the computer,
application of the FEM was limited to small scale A B
laboratory models. On the other hand, the need for the
theoretical prediction of the welding distortion and residual
stress, especially in the industries, has kept growing. To
meet this demand, authors developed an FE code which is
specially designed for efficient computation so that
complex and large welded structures or welded parts
produced in industry can be simulated. In this paper, the
idea of iterative substructure method (ISM) [1-5] and its
advantage for large scale problems are discussed. Also, its A B’
ǻ
potential capability is demonstrated through an example
problem with more than million elements.
B
2. Concept of Iterative Substructure Method
If the problem can be separated into a large but constant Fig. 1 Concept of Iterative Substructure Method.
stiffness problem and a small moving nonlinear problem,
the computational time can be reduced by using an iterative
substructure method. As schematically shown in Fig. 1, the
region B which exhibits strong nonlinearity is limited in a
Solve A using stiffness of A’ and save matrix
small region compared to the whole model A to be analyzed. after forward elimination
In ISM, the problem is reformulated as follow.
(1) The region B is the region with strong nonlinearity.
(2) The region (A-B) is the mostly linear region. Extract displacement uǻ on ǻ
(3) The boundary 䂳 is the boundary between regions
(A-B) and B. Solve B under boundary displacement uǻ
(4) The virtual region A’ is the model in the past. The
difference from region A is that its stiffness is
unchanged until updated. Compute unbalance force on ǻ
(5) Solve the region (A-B) using the stiffness of A’ and {A̓-B}+B
solve the region B using current stiffness.
(6) The continuity on the boundary 䂳 is maintained
through the iterative procedure. Convergence check
79
Thermal elastic plastic analysis for welding problem of large scale models
Thickness : 2mm
Welding Line
㻡㻌㻝㻜㻡
Fig. 5 Geometry of weld joint with cross-section.
Thermal analysis
Mechanical analysis
㻡
㻠㻌㻝㻜
Computong time [s]
㻟㻌㻝㻜㻡
㻞㻌㻝㻜㻡
㻝㻌㻝㻜㻡
Number of elements and the model with 1,208,000 elements can be computed in
Fig. 4 Relation between computing time and number of 434,010 seconds (120.5 hours).
elements for thermal and mechanical analysis.
4. Preliminary Computed Results for NRC Round
displacements in the whole model. If the updated
Robin
stiffness of A is used, save the stiffness matrix after
the forward elimination. The second example is the preliminary computation of
(b) Solve B by using the displacement on 䂳 computed in the welding residual stress in the nozzle shown in Fig. 5.
step (a) as the boundary condition. In this way, the This model is proposed by US Nuclear Regulatory
continuity of the displacement on the boundary 䂳 is Commission for International Weld Residual Stress Round
maintained. This solution process is nonlinear and the Robin [6]. As a preliminary analysis, the residual stress
Newton-Raphson iterative method is employed. produced by the welding between the nozzle and the safe
(c) Compute the unbalance force between the reaction end was computed. The outer and the inner diameter of the
forces from (A’-B) and B. safe end are 381 mm and 284.5 mm, respectively. This
(d) To recover the continuity of the traction, feedback the model is divided into 63,000 elements. There are 40 and 27
above unbalance force to step (a) and compute the passes on the outer and the inner sides. In the computation,
only the welding passes on the outer and inner surfaces are
correction for the displacement on the boundary 䂳.
computed using moving heat source and a stationary ring
(e) Repeat steps (a) through (d) until the convergence is
shape heat source (equivalent to axi-symmetric model) is
reached.
used for the rest of the passes. For this computation, the
total computing time for the mechanical analysis is 53.6
3. Capability of ISM for Large Scale Welding Problems hours. The distribution of the axial component of the
The model is a simple bead on plate welding model with residual stress is shown in Fig. 6. Through this preliminary
different width. In all cases, the thickness and the length of computation, it is found that the full computation using a
the plate is fixed as 2 mm and 100 mm, respectively. The moving heat source for all passes can be completed in about
width of the plate is increased from 100 mm to 12,000 mm 8 days.
so that the number of elements is increased from 18,000 to
1,208,000. An example for the case with 18,000 elements is 5. Conclusions
shown in Fig. 3. The relations between computing time and
ISM was developed based on the idea to separate the
the number of elements are shown in Fig. 4. For these
nearly linear region and the nonlinear region in an efficient
computations, Intel Xeon W5590 (Quad Core 3.33GHz,
way. In order to examine the effectiveness of this method,
96GB) was used under single core mode. According to
the bead on plate welding was analyzed as example. From
Fig.4, the mechanical analysis of the model with 18,000
that computation, it is found that the computational time is
elements can be completed in 6,542 seconds (1.8 hours),
drastically reduced by using ISM and its advantage
becomes greater when the number of elements increases.
80
Transactions of JWRI, Vol.39 (2010), No. 2
81
Transactions of JWRI, Vol.39 (2010), No. 2
†
Comparison on several kinds of T-E-P FEM software for welding
KEY WORDS: (Software Comparison) (T-E-P FEM) (Welding Temperature) (Welding Deformation) (Welding
Stress and strain)
3. Calculated Results
Comparison of temperature
Figure 1 shows the distribution of temperature at
center cross-section on the surface of plate when welding
time is 21 sec, and total heat input parameter (Q/h2) is 4.06
(J/mm3), welding speed is 10mm/Sec.
From Fig.1, it is clearly seen that temperature results
obtained from three kinds of software are consistent, the
peak temperature has difference a bit.
82
Comparison on several kinds of T-E-P FEM software for welding
83
Transactions of JWRI, Vol.39 (2010), No. 2
84
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Finite element simulation) (Multi-pass welding) (Rezoning technique) (Welding Deformation)
㻌 㻌 㻌 㻌 㻌 㻌 㻌 㻌 㻌 㻌 (Residual Stress) (Computational efficiency)
85
Finite element simulation of multi-pass welding process with rezoning technique
86
Transactions of JWRI, Vol.39 (2010), No. 2
4. Conclusions References
Multi-pass butt joint welding process is simulated by [1] L. E. Lindgren, H. A. Haggblad, J. M. McDillb: a.s.oddy,
three dimensional FEM in both rezoning model and full Comput. Methods Appl. Mech. Engrg. 143(3)(1997), pp.401-
models. Transient temperature, welding deformation and 409.
residual stress calculated in rezoning model agree well with [2] H. Murakawa, I. Oda, S. Ito, H. Serizawa, M. Shibahara and
that of full model. Computational time of the simulation is H. Nishikawa: J. Kansai Soc. N. A. Japan, 243(2005), pp.67-
reduced by 36% with a rezoning technique for the studied 7 0 .
model. Effectiveness and efficiency of the proposed [3] S. B. Brown and H. Song: ASME J. Engrg Industry,
method is confirmed. 155(1993), pp.415-423.
87
Transactions of JWRI, Vol.39 (2010), No. 2
WANG Jiangchao *, MA Ninshu **, MURAKAWA Hidekazu **, TENG Bugang ***
and YUAN Shijian ***
KEY WORDS: (Welding Distortion) (Inherent Deformation) (Shell Structure) (Elastic Analysis)
(Thermal-Elastic-Plastic computation) (Finite Element Method)
1. Introduction
In the manufacturing industry, welding is commonly
used as a main joining process to assemble various parts to
a welded structure. However, welding distortion always
occurs during the welding process because of non-uniform C D 2
expansion and contraction caused by the heating and A
B E 1
cooling cycle at the weld and surrounding base material [1].
Distortion due to welding not only degrades the Top polar Bottom polar
performance of a welded structure but also increases its
production cost. In fact, it is impossible to completely
eliminate welding-induced distortion, but a welded structure
can be produced within a small tolerance through Fig. 2 Position of measuring points
minimizing the welding-induced distortion [2]. Therefore,
prediction and control of welding distortion have become of 3. Computational Approach
critical importance during assembly processes. The computational approach for welding distortion of
the spherical structure is divided into two steps. In the first
2. Experimental procedure step, the inherent deformations in a butt welded joint are
The research object is a spherical structure which estimated by a three dimensional thermal-elastic-plastic
consists of 16 bent thin shells and 2 circular polar plates. FEM. Then, in the second step, welding distortion induced
The spherical structure has been pre-assembled by tack by sequential welding of the spherical structure is computed
welds. The dimensions of this spherical structure are shown by an elastic FEM in which the inherent deformations
in Fig. 1. To evaluate welding distortion, five typical points predicted in the first step are used as the input data.
as shown in Fig. 2 have been selected to measure the
displacements after each welding line. [1] Estimation of inherent deformation using thermal-
elastic-plastic FEM.
The dimensions of a butt welded joint are 400mm×
400mm×1mm, shown in Fig. 3. The base metal is a low
carbon steel and the welding condition is shown in Table. 1.
88
Prediction and measurement of welding distortion of thin shell structure
-1.0+00
[2]Elastic analysis with inherent deformation -2.0+00
Based on the inherent strain theory, welding induced
-3.0+00
distortion of the spherical structure can be calculated using
the known inherent deformations. Part numbers of the -4.0+00 Z
spherical structure are shown in Fig. 5. The welding -5.0+00 X
sequence is corresponding to the part member. Y
Component Max Min
Y-Direction Displacement 4.7461mm -4.5452mm
Fig. 6 Welding distortion of spherical structure
89
Transactions of JWRI, Vol.39 (2010), No. 2
㻞 4. Conclusions
(1) Welding distortions of a spherical welded structure
㻝
have been predicted by an elastic FEM using the
㼃㼑㼘㼐㼕㼚㼓㻌㻰㼕㼟㼠㼛㼞㼠㼕㼛㼚㻌㻔㼙㼙㻕
㻙㻝
㻙㻞
㻙㻟
㻯㼛㼙㼜㼡㼠㼍㼠㼕㼛㼚㼍㼘㻌㼞㼑㼟㼡㼘㼠㻌㼛㼒㻌㻼㼛㼕㼚㼠㻌㻭
㻙㻠 㻹㼑㼍㼟㼡㼞㼑㼐㻌㼞㼑㼟㼡㼘㼠㻌㼛㼒㻌㻼㼛㼕㼚㼠㻌㻭
㻯㼛㼙㼜㼡㼠㼍㼠㼕㼛㼚㼍㼘㻌㼞㼑㼟㼡㼘㼠㻌㼛㼒㻌㻼㼛㼕㼚㼠㻌㻯
㻹㼑㼍㼟㼡㼞㼑㼐㻌㼞㼑㼟㼡㼘㼠㻌㼛㼒㻌㻼㼛㼕㼚㼠㻌㻯
㻙㻡
㻜 㻞 㻠 㻢 㻤 㻝㻜 㻝㻞 㻝㻠 㻝㻢
㼃㼑㼘㼐㼕㼚㼓㻌㻼㼍㼟㼟
Fig. 8 Comparison between computational and measured
result of point A and C
90
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Ductility dip crack) (Welding) (Inconel600) (Finite element method)
1. Introduction
The Ni-base alloys have widely been used in nuclear
power plant, particularly to variety pipe welding.
Ductility Dip Crack mainly happens in a Ductility Dip
Temperature Range (DTR) below the solidus temperature
when a stroke is exerted[1]. Currently, among all the
Ni-base alloys that have been used, Inconel600 is a kind
of Ni-Cr-Fe alloy with excellent high temperature
strength and corrosion resistance. A significant amount of (a) Welding orientation ˄b˅Shape of samples
research has been performed to identify the mechanisms Fig. 1 Schematic diagram of sample
of DDC. Rines and Wray’s [2] work shows that DDC is 2
formed by void accumulation during shear loading, Samples are heated to 1000 & and held for 10s. And
which has great similarity to creep cracking. Another then cooled to a certain temperature (750ćˈ850ćˈ
hypothesis is studied by Zhang, et al. [3], who believe that 900ćˈ950ćˈ1000ćˈ1050ć), a stroke is loaded after
intergranular precipitate, intergranular sliding and holding the temperature for 10s. The stroke is also held
boundary tortuosity have complex effects on DDC for 10s, later, cooling till room temperature. Along with
susceptibility. And K. C. WU and R. E. [4] Herfert think the surfaces holding at 1100ć, 1150ć, 1200 ć, the
that appearance of DDC is due to grain boundary critical strain of DDC is found.
weakening, as intergranular strength be comes lower then Finally, corroding on the sample surfaces, optical
intragranular strength when heated up to a certain microscope and SEM are used to observe the surface
temperature. This work mainly studies DDC cracks and microstructure.
susceptibility of Filler Metal 82 under different stress 1200
states.
1000
91
Investigation on ductility dip cracking susceptibility of filler metal 82 in welding
Strain
Critical strain of DDC
temperature above 1100ć, the strengthened 700 800 900 1000 1100 1200 1300 1400
which reduces the DDC sensitivity effectively. [5] (a) Welding speed of 2mm/s
According to STF results, from 900Υto 1200ćwith 0.16
Equivalent strain
3mm
4mm
0.08
0.14 5mm
6mm
Critical strain of DDC 0.06 Critical strain of DDC
0.12
0.04
Equivalent strain
0.10
0.02
0.08
0.00
0.06
700 800 900 1000 1100 1200 1300 1400 1500
0.04 Temperature / oC
0.02
700 800 900 1000 1100 1200
(b) Welding speed of 5mm/s
o
Temperature / C Fig. 5 Equivalent strain changing with temperature
Fig. 3 Critical strain at different temperature
gets across the sensitive range. It means that the strain at
this location is higher than critical strain and prone to
3. DDC susceptibility under complex stress state DDC propagation.
FEM model of TIG welding is built up by ABAQUS When raising the welding speed in the case of the
common software with size of 100×100×2mm plate. The same weld width, in light of conductive time, strain
welding parameter used are as follows, current I=100A, changes more behind the temperature peak compared to
voltage U=10V, welding speed Ȟ=2mm/s, and a double speed of 2mm/s. Thus deformation is slight in DTR.
ellipsoid heat source is used. Figure 4 shows the From the curve of equivalent strain increment to
distribution of welding temperature. temperature (Fig. 5b), the position around the weld has
less DDC susceptibility.
According to the simulation results of single-pass
welding, welding strain decreases against distance to the
fusion line. Moreover, optimizing welding parameters,
such as increasing welding speed, can effectively
decrease DDC susceptibility. With the increase of
welding speed, strain changes respectively behind the
temperature peak, which enables deformation to avoid
the sensitive temperature range.
92
Transactions of JWRI, Vol.39 (2010), No. 2
tensile test. When deformation is over the critical strain [1]Tang Zhengzhu, Chen PeiYing and Wu Wei. Influence
GB can endure, void accumulation results in crack Mechanism of Nb on Ni-based alloy DDC Susceptibility.
propagation perpendicular to loading. Transactions of the China Welding Institution, 2008, 29(1):
109-112. (in Chinese)
˄2˅By the FEM model of single pass welding, 3mm
[2]Rhinestone, F. N. and P. J. Wray. Investigation of the
from weld center has the most susceptibility at this Intermediate Temperature Ductility-Dip Cracking in
welding condition. Upon the same weld width, Austenitic Stainless Steels and Ni-base Alloys. Welding in
simulation shows that increasing welding speed can the World, 2002. 46(7/8): 32-40.
effectively improve DDC resistance. [3]Zhang. Y. C, Natsuda. H, Matsuda. F. Weldability of
Fe-36%Ni alloy (Report III). Transactions of JWRI, 1985,
14(1): 107-114.
Acknowledgement [4]K.C.Wu and R.E. Herfert. Microstructural Studies of Rense
This project is support by the National Natural Science 41 Simulated Weld Heat-Affected Zone. Welding J,
Foundation of China (Grant No. 50975176) 1967(46):32.
[5]BY F. F. Noecker, J. N. Dupont. Metallurgical Investigation
into Ductility Dip Cracking in Ni-Base Alloys: Part II.
References Welding Research, 2009, 88(3): 62-77.
93
Transactions of JWRI, Vol.39 (2010), No. 2
1. Introduction
Double bottom is a typical unit of hull construction, and This double bottom block is divided into several
the prediction of its welding deformation is essential for components, such as the inner bottom plate, floor plates,
hull structural design and accuracy control of shipbuilding. shell plate, longitudinal girder, which are assembled
Three-dimensional thermal elastic-plastic FEM can separately at five steps as below, and shown in Fig. 1.
simulate the whole welding process, but it takes huge inner bottom plate => longitudinal structure => floor plate
calculation time and is only applicable to welding and stiffener => straight shell plate => curved shell plate.
mechanics behavior analysis of welded joint or small
structures. In this paper, through a large amount of
computation and measurement, a simple calculation formula 3. The Finite Element Model
The finite element model is established by software
of inherent deformation which relies on thickness of plates
HyperMesh. According to symmetry, a half of the
is proposed. Based on the inherent deformation calculated
block structure is meshed with quadrilateral elastic
by the simple formula [1][2], the deformations of the double
shell element [3]. There are 1730 model nodes and 2108
bottom structure in NO.6 Cargo of 50000 DWT
elements. Figure 2 depicts the finite element model
multipurpose ship are predicted by special software WSDP.
and constraints. X axis (longitudinal) is along the bow
The feasibility of inherent deformation database and simple
direction, Y axis (transverse) is along the port
formula is verified for the numerical results and agrees well
direction of the ship and Z axis (vertical) is along the
with the experimental data.
upward direction to the shell plate. 151 fillet welded
2. Production Process of the Double Bottom Structure joints were employed.
The 50000 DWT multipurpose ship is a kind of welded
steel ship with double shell structure. In this paper the block
1C14P/C/S has been taken as specimen of study. The
specimen’s width is 32250mm, length is 11860mm, and the
height is 1750mm. The steel-type of inner bottom plate is
GLDH32, and inner bottom longitudinal is GLAH32 while
in other area is GL-A.
94
Prediction of welding deformation for double bottom structure in cargo hold of 50000 DWT multipurpose ship
3
5. Calculated Results
Figure 3 displays the distribution of welding 0
deformation along the longitudinal and transverse direction, 0 1 2 3 4 5 6 7 8 9 10 11
Along the ship length, x (m)
and the deformation enlarged 100 times, the dashed line is
the original contours of the structure. (a) Along the ship length
21
Figure 4 expresses the shrinkage curve along the Step 1
12
0
0 2 4 6 8 10 12 14
Along the ship width, x (m)
(b) Along the ship width
Fig. 4 Comparison of experimental and computational
displacement by five steps
95
Transactions of JWRI, Vol.39 (2010), No. 2
96
Transactions of JWRI, Vol.39 (2010), No. 2
LUO Jian*
KEY WORDS: (Thermal mechanical property) (GMAW) (Magnetic field) (Overlaying welding)
(Microstructure)
1. Introduction
Hot forge moulding has several remarkable advantages,
including high working performances, high precision, high
wear resistance (especially the red hardness) and good
ductile plasticity, the bimetal material used in hot forge
mould is subjected to the rigorous working environment on
the hot forge process. The GMA Welding with a
longitudinal electromagnetic field (LMF-GMAW) is one of
the overlaying welding methods, which is of low cost and
the high performance to manufacture the bimetal thermal
forming mould. However, there is still a lack of practical
detailed understanding of the thermal mechanical property
about the LMF-GMAW work pieces.
In this paper, we focus on the need for understanding Fig. 1 Diagram of the LMF- GMAW overlaying system
the thermal mechanical property of these work pieces with
overlaying hard alloy components on the carbon steel base
by using the LMF-GMAW method. 㸦 d㸧 㸦 b㸧
97
Microstructure and mechanical properties of overlaying specimens in GMAW hybrid an additional longitudinal
electromagnetic field
98
Transactions of JWRI, Vol.39 (2010), No. 2
4. Conclusions
Based on the comparison study of microstructure and
mechanical properties between LMF-GMAW specimen and
general GMAW, we conclude that LMF overlaying method
can increase surface hardness, reduce surface friction loss,
improve wear resistance, and enhance the thermal physical
mechanical property of specimens.
References
[1] Luo J, Lin Y, Luo Q and Wu YX, Sci. Technol. Weld. Join.,
(a) LMF-GMAW work piece 9(2004),pp 465-469.
99
Transactions of JWRI, Vol.39 (2010), No. 2
LUO Jian*
KEY WORDS: (Welding pool) (GTAW) (Magnetic field) (Flowing behaviors) (Different parameters)
1. Introduction (a (b
The electromagnetic field hybrid arc welding technique
is applied in many manufacturing fields, however there is a
lack of detailed understanding of flowing behavior of t=1.0s t=1.0s
molten metal at a weld pool during GTA welding hybrid of
longitudinal electromagnetic field (LMF-GTAW). For
example, the impact of longitudinal electromagnetic field to
the flowing behaviors of weld pool is not yet clear. Because
the hybrid electromagnetic field can not only affect the
behavior of the welding arc, but also change the driving
force and flowing of molten metal at the weld pool.
In this paper, we focus on the need for understanding
weld pool’s flowing behavior affected by different
parameters and multi-materials in LMF-GTAW using a Fig. 1 The flowing at AlCu4SiMg aluminum alloy welding
numerical simulation research method. pool: (a) GTAW; (b) LMF-GTAW.
Figure 1 shows that the LMF quells the molten metal 3. Results and Discussion
motion on the depth direction (Z) and increases the radial Figure 2 shows the surface tension coefficient can
motion(R), the velocity on the surface of weld pool and the change the flowing direction, the large surface tension
width. Same conditions are carried out in GTAW and LMF- coefficient (absolute value) makes the radius of the stable
GTAW. circular loop decrease. The centre of circular loop has also
a general exodus, the flowing in homogeneity increases and
the speednear the weld pool’s surface is higher than that of
† Received on 30 Sep. 2010 Transactions of JWRI is published by Joining and Welding
* The State Key Lab of Mechanical Transimission, Chonging Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
University, China. Japan
100
Flowing behaviors affected by different parameters and multi-materials in GTA weld pool hybrid a longitudinal
electromagnetic field
other parts that are associated with surface tension (b)
(a) t=1.0s
coefficient absolute value.
Although a strong hybrid electromagnetic field can
make a strong stirring motion, the increased magnetic t=1.0s
induction has little change on the flowing mode or shape on
the ROZ cross-section of weld pool. But the flowing
velocity of molten metal is held in the ROZ plane, as shown
in Fig. 3. The strong stirring behavior is one of reasons.
Undoubtedly, the increased welding current can lead to
fast flowing, shown in Fig. 4. It is also contribute to the
stable flowing path and uniform circular loop. The “step-
style” about the flowing loop disappears at the developing
period of the weld pool. A stable even flowing status is
formed at the welding pool with the big welding current, as
a result of higher temperature.
Fig. 5 Flowing at STB42 carbon steel LMF Weld pool:
(a) (b)
(a) ROZ cross-section; (b) stirring rotation motion.
(a) (b)
t=1.4s t=1.4s
t=0.2s t=0.2s
101
Transactions of JWRI, Vol.39 (2010), No. 2
102
Transactions of JWRI, Vol.39 (2010), No. 2
103
Concept of inherent strain, inherent stress, inherent deformation and inherent force for prediction of welding
distortion and residual stress
taken into account through the standard FE formulation.
G y* bH *y That associated with the inherent force FT must be also
taken into account in the force vector and the stiffness
matrix in FEM.
In case of large deformation problem, the strain and the
G x* aH x* total potential energy of the structure at time t 't can be
symbolically described as the following equations.
H (u 'u ) H (u ) '1H '2H (15)
Fig. 2 Shrinkage represented by inherent deformation.
S u 'u S u '1S '2S (16)
6. Inherent Force in Thin Plate where, u and 'u are the displacement and its increment.
Further, the inherent forces can be defined by the '1H , '2H , '1S , '2S are the first and second order terms of
following equations. 'u in the strain and the total potential energy, respectively.
From the first order term '1S , the force vector is derived
Transverse inherent force FT* E ³ H x* dxdz EhG x* (11)
Longitudinal inherent force FL* E H *y dxdz EhG y* (12) while the stiffness matrix is derived from the second order
³ term '2S . Noting that the inherent force (Tendon force) is
³
Transverse inherent moment M T* E zH x* dxdz EIT x* (13)
the concentrated stress described as a delta function, the
Longitudinal inherent moment M L* E ³ zH *y dxdz EIT y* (14) contributions of the tendon force to the large deformation
through the force vector and the stiffness matrix are given
by the following equations.
7. Procedure to Introduce Welding Inherent Strain to
FEM '1S ..... ³ V (u )'1H dv .. ³ V (u )T (u )'T (u )dv
Since the restraint in the welding direction is strong and .. ³ FT T (u )'T (u )ds (17)
that in the transverse direction is very weak, the inherent
strain in the transverse direction mostly transforms into the
deformation and that in the welding direction transforms Tendon f
into the residual stress. Thus it is natural to use the inherent force
z TFT FT
deformation for the transverse shrinkage and the angular V
distortion. For the longitudinal shrinkage and bending, the T
inherent force or moment must be used as shown in Fig. 3.
In this way, the shrinkage in the weld zone can be described y
by using relatively coarse FE mesh as illustrated in Fig.4. x
8. Large deformation Fig. 5 Stress distribution Fig. 6 Force produced by
To predict the distortion of thin wall welded structures, in element. large deformation.
the large deformation must be taken into account. As shown
in Fig. 5, the distribution of the stress in an element which
is facing the welding line is a combination of that is
distributing in an element and that is concentrated at a point
in the form of delta function. The effect of large
deformation associated with the distributed stress can be
FT ahEH *y
aQH *y QF T / hE
aH *
x Fig. 7 Deflection computed under small deformation.
104
Transactions of JWRI, Vol.39 (2010), No. 2
10. Conclusions
The idea of combined use of inherent deformation and
inherent force is presented and its effectiveness for large
deformation problems is demonstrated.
105
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Inherent strain) (Inverse analysis) (Welding deformation) (Finite Element Analysis)
2. Principle of inverse analysis method for inherent Fig. 1 The principle sketch of the inverse analysis method
strain
The principle flowchart of the developed method is
shown in Figure 1. The measured displacement could be
expressed as the product of elastic response matrix and the
unknown coefficient vector of inherent strain. The elastic
response matrix is related to the welding structure. If the
unknown coefficient vector of the inherent strain is
obtained, the inherent strain can be calculated by
Fig. 2 The plug-in menu of the inverse analysis tool
multiplying the coefficient vector and the distribution
function vector. The distribution function of inherent strain
106
Inherent strain calculation from inverse analysis of measured welding deformation based on python of
ABAQUES CAE
(a) (b)
Fig. 5 Distribution of inherent strain: (a) H x , (b) H y
I I
6.00%
5.00%
Relative Error
4.00%
3.00%
2.00%
1.00%
107
Transactions of JWRI, Vol.39 (2010), No. 2
108
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Inherent Deformation) (Interference Element) (Numerical simulation) (Welding Distortion)
1. Introduction
Fusion welding processes are widely used to assemble both the longitudinal stiffener and the transverse stiffener is
large thin plate structures such as ships, automobiles and 12mm. The size of each mesh is 50mm×50mm. Using this
passenger trains because of the high productivity. However, model, five cases as shown in Table 1 are examined to
welding-induced distortion often inevitably occurs during investigate how the gap and misalignment generate during
the assembly process. Welding distortion not only reduces the assembly, and how welding sequence and gap
fabrication accuracy of a welded structure but also correction affect the final distortion. It is assumed that the
decreases productivity due to correction works. If welding fillet joints and cross-shaped in model A are performed by a
distortion can be predicted using a practical method single-sided welding process. The welding conditions of
beforehand, the prediction will be helpful for taking each joint are the same as Ref.[1].
appropriate measures to control the dimension accuracy. In
this study, an elastic FEM approach to welding distortion Table 1 Simulation cases of Model A
for large and complex welded structures is developed based
on inherent strain theory and interface element formulation Case Detailed information
[1-3]. Welding distortions in two typical structure models A0 Weld all parts at the same time
(model A and model B) are simulated using the proposed
method. Model A is employed to show how the gap and A1 1) Weld transverses/longitudinal stiffeners
misalignment generate during the assembly. Model B is a 2) Correct gaps 3) Weld stiffeners and skin plate
large thin plate structure, and it is used to compare the 1) Weld long. Stiffeners and skin plate 2) Strongly
welding distortions computed by large- deformation theory A2 correct the gaps between trans. stiffeners and skin
and small-deformation theory. plate 3) Weld transverse stiffeners and skin plate 4)
Weld longitudinal stiffeners and Trans. stiffeners.
2. Finite Element Models
1) Weld long. Stiffeners and skin plate 2) Weakly
In this study, a two-step computational procedure [1] is
used to simulate welding distortion occurring during the A3 correct the gaps between trans. stiffeners and skin
assembly process. At the first step, the inherent plate 3) Weld transverse stiffeners and skin plate 4)
deformations of each type joint are estimated by thermal Weld longitudinal stiffeners and transverse
elastic plastic FEM using a small model. In the thermal stiffeners.
elastic plastic FE analysis, the detailed welding conditions 1) Weld Trans. Stiffeners and skin plate 2) Strongly
and temperature dependent thermal physical/mechanical A4 correct the gaps between longitudinal stiffeners and
properties are considered. At the second step, the total skin plate 3) Weld longitudinal stiffeners and skin
distortion of the large model is estimated by the developed plate 4) Weld longitudinal stiffeners and transverse
elastic FEM based on inherent strain theory and interface stiffeners.
element method. In this step, the welding distortion induced
by heat input is simulated by inherent deformation, while
the formation of gap/misalignment, the welding sequence
and the gap correction is modeled by interface elements.
Figure 1 shows the mesh division and boundary
conditions of Model A. This model consists of one skin
plate, two longitudinal stiffeners and two transverse
stiffeners. The length, width and thickness of skin plate are
4000mm, 2000mm and 9mm, respectively. The thickness of
109
Application of inherent deformation and interface element to prediction of welding distortion during assembly process
Welding line
4000mm
2000mm
GAP
Misalignment
Fig. 1 Finite element model A (Model A)
5
Line 1
0
12000mm
Deflection (mm)
-5
4000mm
-10 Case A-0
CaseA-1
CaseA-2
Fig. 2 Finite element model B (Model B) -15
CaseA-3
CaseA-4
conditions of model B. The length of skin plate is 0 500 1000 1500 2000
X-Coordinate (mm)
2500 3000 3500 4000
110
Transactions of JWRI, Vol.39 (2010), No. 2
Deflection distributions along line 1 respect to the center- line. Whereas, the deflection along
30
Case B-1 (Large)
line 1 of case B-2 has a symmetrical distribution.
2 (Small)
Case B-1
25
Asymme 4. Conclusions
trical
20
deflectio 1) During the assembly process, besides welding process,
welding sequence and gap correction have a significant
Deflection (mm)
n
15
influence on the final deformation.
10 2) For large thin welded structures, because of the
5
propensity of buckling distortion, it is necessary
consider geometrical nonlinear problem.
0
References
0 2000 4000 6000 8000 10000 12000
-5
X-Coordinate (mm) [1] D. Deng and H. Murakawa, Comput. Methods Appl. Mech.
Engrg. 196, 4613-4627,2007.
Fig. 6 Deflection distributions along line 1 (Model B) [2] H. Murakawa, D. Deng,et.al, Trans. JWRI, Vol.38, No2, 63-69,
2009.
significantly larger than that of Case B-2. Moreover,
because of buckling distortion produced in case B-1, the [3] D. Deng and H. Murakawa, Comput. Mater. Sci, 43, 519-603,
2008.
deflection along line l has an asymmetrical distribution with
111
Transactions of JWRI, Vol.39 (2010), No. 2
new JWELD
Simulation for multi-welding lines
Design Welding
Structure Structure Assemble Optimization
Experiment
old
Fig. 1 Simulation stage for a welded structure
JVISION/Post
Comparing with other commercial CAE software which (1) Visualization
(2) Deformation
is mainly applied to predict the welding deformation for a
(3) Export interface
certain welded joint, JWELD is a dedicated and integrated
simulation system which is available not only for a simple Fig. 2 Schematic showing of simulation system JWELD
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* JSOL Corp. Osaka, Japan Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
** JWRI, Osaka University, Osaka, Japan Japan
112
Development of simulation system JWELD for assemble deformation of welded structures
part-
3/t= Tack
6 we ld
s
If a geometry model of a welded structure is built and
the assembly sequences are given, the deformation due to
weld line fitting, tack welding and final welding can be
part
predicted by FEM solver JWELD within very short CPU - 1/t=
6
time. Furthermore, in order to control the deformation, the Part-2
/t4.5 W1
optimization of assemble sequences is being developed and
will be released for next version.
The post processor JVISION/Post can directly show Fig. 4 Deformation due to a single-sided fillet welding
the deformation produced by each weld line. The geometry for a H-section joint
with assembly welding deformation can be exported for the
other purposes of simulation.
Table 1 Arc welding conditions and inherent deformation Fig. 5 Correction deformation of upper flange modeled by
I U v Gx Gy Tx Ty the interface element formulation
(A) (V) (mm/s) [mm] [mm] [rad] [rad]
180 24.0 5.0 0.061 0.11 0.00 0.021 4. Applications of Simulation System JWELD
(1) Welding assembly deformation in a grid structure
The single sided weld line of fillet joint is located on the Figure 6 shows a long grid structure which has ten
right side of flange. Therefore, the deflection on the right parts and nine welding lines. The welding assembly
side of flange is larger than that on the left side. sequences are also shown in Fig.6. The grid made from
steel SS410 is assembled by arc welding process. The
welding current, voltage and welding velocity are 400A,
0 20V, 10mm/sec for three butt weld lines (W1, W2 and W3)
20
50 and 250A, 25V, 10mm/sec for fillet weld lines (W4, W5)
t=6
.5 and (W6, W7, W8, W9), respectively. The welding lines
t=4 line
W4, W5 are double sided fillet welds and weld lines W6,
eld
W W7, W8, W9 are single sided fillet welds. The inherent
400 deformations of these weld lines are predicted using a data
base creating program JWDB. Figure 7 shows the
computed final deformation due to sequential assembly
welding from W1 to W9.
Fig. 3 Deformation due to a single-sided fillet welding (2) Welding assemble deformation in a train car model
Figure 8 shows a simplified FEM model [5] of a train
Figure 4 shows the welding deformation of H type car structure for the computation of welding deformation.
section of fillet welds due to a single-sided weld line (W1) The details are neglected because the limitation of space.
between the lower flange (part-1/t=6) and web (part-
2/t=4.5). If two tack welds between web and upper flange Length=10,000mm part6
(part-3/t=6), the angular distortion due to the weld line Width=1,000mm
part5 W3
(W1) will make the upper flange deform following the web. Height=1,000mm
Thickness=10mm
If there are no any tack welds between the web and upper W5 part4
W8
W2
flange, the gap and misalignment at the interface will part3 W9
produced. If correction of the gap and misalignment in both W6 W4 W1 part2
part9
part10
part1
transverse direction and normal direction are modeled by
W7
the interface element formulation [2] between the upper
part7 part8
flange and the web, the upper flange will deform only in y
direction and z direction as shown in Fig. 5 by two arrows.
Fig. 6 Gird model for computation of welding deformation
113
Transactions of JWRI, Vol.39 (2010), No. 2
5. Summary
(1) Integrated simulation system JWELD was developed
for industrial users.
(2) The validity of this system was verified by basic
welded joint models and structural models.
Reference
[1] Y. Ueda, H. Murakawa and N. Ma: Computational Approach
to Welding Deformation and Residual Stress, Sanpo
publication, ISBN978-4-88318-034-9 C3057, 2007
[2] D. Deng and H. Murakawa: FEM Prediction of Buckling
Distortion Induced by Welding in Thin Plate Panel Structures,
Computational Material Science, 41 (2008) 591-607.
[3] JSOL: JVISION User’s Manual, //www.jsol.co.jp/cae/
Fig. 7 Computed welding deformation of gird model [4] JWES: //www-it.jwes.or.jp/nedo01
[5] T. Hisamori and E. Tachibana: Numerical Simulations of A
Train Collision with A Building Structure, Japan Architecture
Society, 2006, p.43-52, http://jweld.jsol.co.jp/
2,550
00
20,0
3,00
0
114
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Martensite Transformation) (Retained Austenite) (In situ Observation) (X-ray Diffraction)
(Synchrotron Radiation)
115
In-situ Observation of martensite transformation and retained austenite in supermartensitic stainless steel
116
Transactions of JWRI, Vol.39 (2010), No. 2
117
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Laser welding) (Weld metal) (In-situ observation) (Nucleation site)
118
In-situ observation of phase transformation during laser welding processes
was observed.
Figure 3 shows the HAADF-STEM image, EDS
mapping analysis, EDS spectrum and selected area
diffraction pattern from inclusions in Y2 in which acicular
ferrite was observed. This inclusion was surrounded by a
titanium enriched layer. This layer was identified as TiO by
EDS spectrum and SAD. The TiO has a Baker-Nutting
orientation with acicular ferrite. In this case, the lattice
misfit between TiO and acicular ferrite is 3.0%[2].
Therefore, the TiO on the inclusion surface promotes
intragranular nucleation supplying low interface energy.
Figure 4 shows the corresponding crystal-orientation
maps measured in the Y2(Al/O=0.73) by EBSD analysis.
Inverse pole figure color maps show the distributions of
ferrite and austenite, respectively. The pole figure of ferrite Fig. 3 TEM analysis of inclusion after thermal cycle in
shows {110} and {111}. Moreover, the pole figure of Y2(Al/O=0.73). (a)HAADF-STEM image ; (b)EDS
austenite shows {111} and {101}. From these pole figure, mapping by Ti-K Ș X-ray ; (c)EDS spectrum from the point
{110} ȘFe and {111} Ț Fe , {111} ȘFe and {101} ȚFe were indicated by a cross in the HADDF image; (d) Selected area
parallel, respectively. Therefore, these results suggest that diffraction pattern from the area indicated by a circle in the
they have Kurdjumov-Sachs (K-S) orientation HADDF image and its key diagram.
relationship.
Figure 5 shows the schematic illustration for
orientation relationship between TiO and acicular ferrite,
orientation relationship between acicular ferrite and
austenite. The inclusions which related to acicular ferrite
formation were surrounded by TiO. The B-N orientation
relationship was satisfied between the TiO and acicular
ferrite. The K-S orientation relationship was satisfied
between the acicular ferrite and austenite. Therefore, a
crystal plane of acicular ferrite corresponds to the habit
plane {111} ȚFe and the growth direction was <111> ȘFe .
119
Transactions of JWRI, Vol.39 (2010), No. 2
120
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: &reep damage1umerical simulationContinuum damage mechanicsType IV cracking
ASMEP92 steel
dt
3
2
n 1
BV e S ij 1 U U 1 D
n
(1)
(DV 1 (1 D )V e )
v
dD A
g (2)
1 D
I
dt I 1
(1/ ( I 1))
Dcr 1 (1 g ) (3)
3. FEM model
The welded joint of P92 steel analyzed in this paper is
kept at 650 °C and 70 MPa.The creep specimen of a welded
joint where four different material properties namely BM,
CGHAZ, FGHAZ, and WM are taken into account is a
round bar. The bar has a cross section of 10 mm and a
gauge length of 50 mm with a HAZ width at 2 mm Fig. 2 Contours of creep damage in weld joint with various
(FGHAZ width 1mm, CGHAZ width 1mm) and groove creep times
121
Numerical simulation on type IV cracking of ASME P92 steel at high temperature
122
Transactions of JWRI, Vol.39 (2010), No. 2
5. Conclusions References
(1) The creep damage in FGHAZ is severe and the failure [1] C. Petry, and G. Lindet: Modelling creep behaviour and
life of FGHAZ is 7100 h. The stress triaxiality and failure of 9Cr-0.5Mo-1.8W-VNb steel, Int. J. Pres. Ves. Pip.,
maximum principle stress located at the FGHAZ are 86-8 (2009), 486-494.
higher than that of other zones in the welded joint [2] J.A. Francis, W. Mazur, and H.K.D.H. Bhadeshia: Type IV
during creep. It implies that they can accelerate the cracking in ferritic power plant steels, Mater. Sci. Technol.,
creep strain development and creep damage 22-12 (2006), 1387-1395.
accumulation; as a result Type IV cracking is likely to [3] S.-T. Tu, P. Segle, and J.-M. Gong: Creep damage and
occur in this zone. fracture of weldments at high temperature, Int. J. Pres. Ves.
(2) A small groove angle and a narrow HAZ width of the Pip., 81-2 (2004), 199-209.
welded joint are calculated to decrease the maximum
principle stress and stress triaxiality in the FGHAZ,
which leads to a decline in the accumulation of creep
damage and can suppress Type IV cracking in ASME
P92 steel.
123
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Triplex stainless steel) (Retained austenite) (Kinetics) (Solidification mode) (Segregation)
(Numerical simulation) (Martensitic transformation) (Ferritic-austenitic transformation)
124
Numerical simulation of austenite retention in triplex stainless steel weld metals
125
Transactions of JWRI, Vol.39 (2010), No. 2
126
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Hardness prediction) (Temper bead welding) (HAZ) (Neural network) (FEM)
127
Prediction of hardness in HAZ of low-alloy steel produced by temper bead welding using neural network
(a)
+Y
,62KHDWWUHDWPHQW
7KHUPDO&\FOHSDVV
7KHUPDO&\FOHSDVV
(b)
7&73
Fig.1 Relationship between TCTP and Hardness of multi-pass
thermal cycle
%0
128
Transactions of JWRI, Vol.39 (2010), No. 2
4. Conclusions
(a) (1) Thermal Cycle Tempering Parameter (TCTP)
calculation method for multi-pass thermal cycles
process has been proposed based on LMP.
(2) On the basis of experimentally obtained hardness
data base and thermal cycle parameters calculated by
FEM, the hardness distribution in HAZ was
predicted using Neural Network.
(3) The predicted hardness was found in good
accordance with the experimental result. It follows
(b)
that the proposed prediction system is effective for
estimating tempering effect in multi-pass welding.
References
[1] J. Liao, K. Ikeuchi, F. Matsuda: Nuclear Engineering and
Design 183 (1998) 9–20.
[2] D. Deng, H. Murakawa: Computational Materials Science
37 (2006) 209–219.
Fig. 5 Comparison between measured and calculated Hv [3] K.G. Samuel, S.K. Ray.: International Journal of Pressure
of (a) 1-layer and (b) 3-layer Vessels and Piping 83 (2006) 405–408.
[4] D. Casasent, X. Chen: Neural Networks 16 (2003) 529–53
129
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Surface tension) (Molten iron) (Oxygen adsorption) (Marangoni convection) (electromagnetic
levitation) (Oscillating droplet method) (High temperature melt) (Undercooling)
130
Surface tension of molten iron measured by oscillating droplet method using electromagnetic levitation
131
Transactions of JWRI, Vol.39 (2010), No. 2
132
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (TA15 titanium alloy) (304 stainless steel) ('ual-pass electron beam welding) (:elding
sequence) (0icrostructure) (Intermetallic layer) (Cluster)
133
Effect of welding sequences on the microstructure of electron beam welded TA15 titanium alloy and 304 stainless
steel joints with copper filller metal
a Intermetallic layer a
,
Crack
TA15 304
ȝP
b
b
Intermetallic layer
II TA15
ȝP
Fig. 1 Macrostructures of the cross sections of the joints
under the two welding sequences. (a- The fisrt sequence, b- Fig. 3 Microstructure in zone I and II in the joints under the
The second sequence) two welding sequences. (a- Zone I in the fisrt sequence, b-
Zone II in the second sequence)
a
the last pass for the first sequence. But for the second
sequence, it is formed in the first pass, so it was reheated by
the second pass. Zone I is characterized by three phases, i.e,
dark gray TiCu, light gray TiFe 2 and solid solution of
copper with yellow color. But zone II only contains light
gray TiFe 2 and solid solution of copper with a yellow color.
Comparing the two joints, there is only little solid solution
of copper in zone I near the intermetallic layer, but more in
zone II. Copper is a soft metal, which can accommodate the
deformation, as a result, the heat stress in the region was
ȝP reduced. The more solid solution of copper there is, the
more prominent the reduction effect is. Consequently, a
crack free joint is obtained under the second sequence. The
b tensile strength is 234MPa and the joint fractured in the
intermetallic layer when it was stretched.
4. Conclusions
(1) A crack free electron beam welding joint of TA15 and
304 stainless steel is obtained by using a copper
sheet as filler metal. One important function of copper
is to reduces heat stress in the weld.
(2) The microstructre and property of the joint is affected
by the welding sequence. High quality joints can be
obtained under the welding sequence in which the
ȝP copper distributed uniformly in the weld.
134
Transactions of JWRI, Vol.39 (2010), No. 2
[2]. X. J. Yuan, G. M. Sheng, B. Qin: Mater Character, 59 [6]. B. Qin, G. M. Sheng, J. W. Huang, B. Zhou: Mater
(2008), pp.930-936. Character, 51 (2006), pp.32-38.
[3]. A. Elrefaey, W. Tillmann: J. Mater Processing Technol, [7]. P. He, X. Yue, J. H. Zhang: Mater Sci and Eng. A., 486
209 (2008), pp.2746-2752. (2008), pp.171 -176.
[8]. Z. Sun, R. Karppi: J. Mater Processing Technol, 59 (1996),
[4]. X. Yue, P. He, J. C. Feng: Mater Character, 59 (2008),
pp.257-267.
pp.1721-1727.
[9]. M. Ghosh, S. Chatterjee: Mater Sci and Technol, 48 (2002),
[5]. A. Fuji, K. Ameyama, T. H. North: J. Mater Sci, 31 (1996),
pp.393-399.
pp.819-827.
135
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Grain size) (Solidification cracking) (Stainless steel) (Laser welding) (Fine grain) (In-situ
observation) (Critical strain)
136
Effect of grain size on solidification cracking susceptibility of type 347 stainless steel during laser welding
4. Conclusions
The effect of columnar grain size on solidification
cracking susceptibility of Type 347 stainless steel was
investigated by using U-type hot cracking test with
developed in-situ observation method. The obtained
conclusions are as follows.
(1) Specimens whose grain sizes of columnar structure in
ZHOGPHWDOYDULHGIURPWRȝPZHUHSUHSDUHGE\
changing the grain size of the base metal using heat
treatment.
(2) Local critical strains of solidification cracking were
measured by four specimens with different grain sizes
RI ZHOG PHWDO IURP WR ȝP +LJK WHPSHUDWXUH
Fig. 3 High temperature ductility curves of used specimens ductility curves were obtained based on various critical
with different grain size strains.
(3) The effect of grain size on solidification cracking
susceptibility was evaluated quantitatively by using the
Figure 4 shows the critical strain rate of temperature
CST of high temperature ductility curves. It was found
drop (CST) based on the high temperature ductility curves
that a specimen of coarse columnar grain size of the
in Fig. 3. The CST of specimen-0h is about 4.7 (×10-4°C-1).
weld metal shows a high solidification cracking
With increasing the grain size, the CST declines
susceptibility comparing with that of fine grain size.
considerably to about 3.8 (×10-4°C-1) for the specimen-0.5h.
Then, the CST drops slowly with increasing the grain size. References
The CST shows about 3.4 (×10-4°C-1) for specimen-4h the [1] Luo X.J.: Study on hot cracking in heat affected zone of
largest grain size used. Therefore, it can be seen that a nickel-base superalloys Inconel718 by laser beam welding. Ph.D.
coarse grain size of base metal induces a high solidification Dissertation of Hiroshima University. Hiroshima Japan.
cracking susceptibility. That is to say, regarding the [2] Nakata K. and Matsuda F. : Quarterly Journal of the JWS,
solidifying weld metal, the larger grain size, the lower 13(1), 1995, 106-115.
137
Transactions of JWRI, Vol.39 (2010), No. 2
138
Transactions of JWRI, Vol.39 (2010), No. 2
KAWAKAMI Hiroshi *, TAMAKI Koreaki **, SUZUKI Jippei *, TAKAHASI Kanta ***,
IMAE Yousuke *** and OGUSU Soichiro ***
KEY WORDS: (Image processing) (Carbide) (SEM image) (Heat affected zone) (2 1/4Cr-1Mo steel) (SR
embrittlement) (Crack propagation) (Crack initiation) (Cleavage fracture) (Facet)
2. Specimen
2 1/4Cr-1Mo steel was used in this experiment. The Fig. 2 Example of image processed image with some
synthetic HAZ specimen was produced by a weld thermal carbide particles on fracture surface.
simulator and tempered at 975K. SR embrittlement occurs
from 10h as shown in Fig. 1. The tempered specimen was
fractured at very low absorbed energy with producing
cleavage fracture. The fracture surface and the cross-section
of each specimen were etched by 3% nital to reveal the
contours of carbide particles.
139
Image processing analysis of growth of carbide particles promoting SR embrittlement in HAZ of 2 1/4Cr-1Mo steel
4. Experimental results
The relationship between aspect ratio and diameter of
major axis of the carbide particle was measured by image
processing as shown in Fig. 4.
The growth direction of the carbide particle is stable
until 50h because gradient in these figures for cross section
area does not change significantly. This result shows that
the shape of the carbide particle retains the needle shape in
short-time SR treatment. Moreover, the carbide grows as a
polygonal type which has small aspect ratio against
diameter of major axis after 100h. In case of 500h, “p” and
“q” in Fig. 4(g) correspond with the polygonal type and the
needle type, respectively. When SR embrittlement occurs
significantly, coarse particles in the cross-section could not
be observed on the fracture surface and aspect ratio on the
fracture surface is lower than the cross-section. This result
shows coarsened carbide particle promote SR
embrittlement as a site of crack initiation or crack
propagation as shown in Fig. 5[3].
5. Conclusions
Image processing was developed to evaluate the shape
of the carbide particle in a SEM image. Detailed analysis
for the shape of the carbide particle by image processing
shows that coarse carbide particles are broken during crack
initiation and crack propagation.
References
[1] H. Kawakami, K. Tamaki, J. Suzuki, Y. Ichikawa.: Q. J.
Japan Welding Soc., 16 (1998), pp.243-251.
[2] K. Tamaki, H. Kawakami, J. Suzuki, T. Niinomi: Q. J. Japan
Welding Soc., 15 (1997), pp.502-508.
[3] H. Kawakami, K. Tamaki, J. Suzuki, K. Takahashi, Y. Imae,
S. Ogusu: Doc. of IIW, IIW-2302-09(2009), pp.1-13.
140
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS˖
˖(Pipeline steel) (Dynamic impact) (Numerical simulation)
1. Introduction Fig.1.
Issues of impact dynamics exist in various
engineering fields. As dynamic fracture problems were
complex in many experiments, people began to simulate
the whole process of impact dynamics and then analyze
the dynamic response of materials under impact loads in a
better way. Nowadays, nonlinear finite element method has
been widely used in various fields.
With the application of dynamic analysis software
ANSYS/LS-DYNA, a model of impact test specimen is
established and a solid model for the simulation impact test
is also set up in which the dynamic impact process for the
impact test is simulated. The absorbed energy is calculated
during the whole impact test and the difference in impact
toughness caused by the notch located in the weld metal,
HAZ and base metal is analyzed. Study proves that the Fig. 1 Three-dimensional numerical model and FEM
numerical model can be used in prediction and evaluation mesh for the impact testing
of the toughness of welded joints. For welded joints, the density is 7820 kg·m-3; elastic
2. Numerical modeling and meshing modulus and tangent modulus is 200GPa and 2.2GPa,
For impact toughness test, finite element software respectively; Poisson ratio is 0.28; hardening exponent is
LS-DYNA is used to carry out the dynamic numerical 0.13; strain rate parameters are 40 and 5 separately.
simulation of X80 pipeline steel welded joint. The actual Yield strength and failure strain have some effect to
impact test process is completely done according to impact toughness. Therefore, the material properties of
Chinese standard GB/T229-1994. In the three-dimensional HAZ (including coarse grain zone and fine grain zone) are
solid model, the supports are simulated as two fixed baffles. simulated by changing these two parameters. The
The dimension of the model is designed by the actual properties of each region in the welded joint specimen are
standard impact specimen. The shape and size of the shown in Table 1. For the impact rammer, the elastic
impact rammer is modeled by the cutting edge shape of the modulus and Poisson ratio is 400 GPa and 0.3, respectively.
actual pendulum. The total width of the HAZ is 3.5mm, The density is 6.8×106 kg·m-3.
among which the coarse grain zone is 1mm and the fine
Table 1 Properties of welded joint at different parts
grain zone is 1.2mm.
The element meshes near the notch are dense while Region Yield strength ı s /MPa Failure strain H
those far from the notch are sparse. When the notch is
Base metal 560 0.5
located in HAZ, weld metal or base metal, the Coarse grain zone 545 0.45
heterogeneity of HAZ has great influence on the specimen Fine grain zone 550 0.47
impact result. The region near the notch is divided to
simulate each area of the HAZ. The properties of material
3. Results and discussion
far away from the notch are set as those nearby. The impact
When the notch is located at weld metal, HAZ and
toughness is simulated respectively when the notch is
base metal, the fracture situation of impact numerical
located at the three different regions. The impact model
model is shown in Fig 2.
graph for the notch is located at weld metal is shown in
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* State key laboratory for mechanical behavior of materials, Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
Xian Jiaotong University, Xian, China Japan
** Tubular Goods Research center of China National
141
Numerical simulation of impact toughness of welded joints for X80 pipeline steel
15000
10000
4. Conclusion
The LS/DYNA finite element software is used to
5000 simulate the impact testing with a reasonable FEM model.
The numerical results show that the impact toughness is
0 affected by yield strength and failure strain. The impact
0.00 0.01 0.02 0.03 toughness is different when the notch is located at different
displacement S/m regions. The impact toughness is the worst for a notch
located at the HAZ and best in the base metal. It can be
Fig. 3 Three force/displacement curves for notch
used to predict the impact toughness of welded joint and
located at different parts provides a new method to evaluate the properties of joint.
Comparing the curves above it can be seen that the
References
location of the notch has a great influence on the impact [1] Atkinson.C, Eshelby JD.:International Journal of Fracture
energy when the difference of material properties between Mechanics, (1968), pp. 3-8.
HAZ and base metal is large. When the notch is at the base [2] Mott NT. Fracture of metals theoretieal consideration.
metal, the force/ displacement curve is the highest and the Engineering, 165 (1948), pp.16-18.
enclosed area is the biggest. This means the impact
toughness there is the best. When the notch is at the weld
metal, the toughness gets worse and becomes worst when
the notch is in the HAZ. That is because the yield strength
and fracture strain have a certain decline. Besides,
softening phenomena in HAZ causes the impact toughness
to reduce significantly.
142
Transactions of JWRI, Vol.39 (2010), No. 2
SUZUKI Shinji *, OYAMA Yusuke *, MAEDA Masakatsu **, and TAKAHASHI Yasuo **
KEY WORDS: (Eco electronics packaging) (Ultrasonic bonding) (Numerical simulation) (Ribbon bonding)
(Power electronics assembly)
1. Introduction the heat input per unit volume and unit time is calculated to
In the recent years, thick Al ribbon bonding has been be 4.0x10-9Jm-3s-1. The heat was supplied only to the
applied to the packaging of power devices such as IGBT elements in contact with the bond-interface in proportion to
(Insulated Gate Bipolar Transistor) modules to secure high the volume of each element.
power density and to improve power control. These are Calculation of plastic deformation of Al ribbon was
very important for green innovations. carried out by using a finite element method[5]. The strain-
Al thick wire bonding is used in the power electronics rate-depending large plastic deformation was taken into
packaging. The wire bonding process has been studied account. The tool and the substrate were assumed to be
experimentally [1, 2, 3] and numerically [4, 5]. These rigid bodies. We assumed that the interface between tool
studies are useful to understand the wire bonding and ribbon was fixed but the interface between ribbon and
mechanism, wire deformation and interfacial behavior substrate was able to slip. The material constants of heat
during bonding. The Al ribbon bonding is affected by conduction and plastic deformation were given from
temperature rise of the bond-area and ribbon deformation. literatures[5, 6].
The temperature rise is related with frictional slip behavior
at the bond-interface. In the present study, a simple 3. Experimental procedure
numerical model of the ultrasonic Al ribbon bonding is The ribbon bonding tests are carried out under the
proposed. The heat conduction between the Al ribbon and condition of ultrasonic power 1~4 W, bonding force
tool (substrate) is modeled to visualize the temperature 1~5.0N and frequency 60kHz. The Al ribbon of 1mm wide
change and distribution. The plastic deformation behavior and Si chip (or SiO2 substrate) were used. The bonding
affected by the friction slip is numerically examined. The time (duration of ultrasonic vibration input) was 100 ~
calculated results are verified, based on the experimental 400ms.
results.
4. Results and discussion
2. Numerical modeling and simulation procedure Figure 2 shows the calculated results of heat conduction
Figure 1 shows the mesh pattern of the bonding tool, Al (change of temperature T with time t at positions A ~ G
ribbon, Al electric pad, and Si chip (or SiO2 substrate). The along y axis of Fig. 1. The bottom of chip and the top of
bonding tool is made of WC. The model is two tool are assumed to be cooled at 300K, i.e., Ts = Tt = 300K,
dimensional and the heat flow in the z direction where Ts is the temperature of imaginary elements (meshes)
(perpendicular to the space) is ignored because the length under the bottom of chip and Tt is the temperature of
of substrate in the z direction is 10 times greater than the imaginary elements on the top of tool. The heat input rate
thickness of the substrate (2.4 mm) and the bottom of the into the bond-interface is constant during bonding. As can
substrate is assumed to be cooled at 300K (in contact with be seen in Fig.1, temperature continously increases at time
a large heat sink). Two boundary conditions were given to t > 0 (after ultrasonic power is inputted) and decreases after
the top of the bonding tool; 1) at 300K and 2) J = 0 in the y t >100s (ultrasonic power is turned off). The temperature
direction, where J is the heat flux. The heat conduction curves for the positions A~F are convex up during bonding
behavior was calculated by using a finite differential but that of position G changes from concave to convex.
method, based on the control volume method. The heat This is due to thermal flow in the x direction. It is
generation at the bond-interface is produced by the friction suggested that Al ribbon heats up to 393K.
slip at the bonding interface. The heat generation is Figure 3 shows the temperature distribution map when
affected by ultrasonic power input. For example, if the heat the substrate is silica and Ts = Tt = 300K without heat
input of 4 W is given to the elements in the vicinity of the resistance of the interfaces. The Al ribbon is heated up
bond-interface and the bonding area is 1mm x 100μm, then uniformly as shown in Fig. 4. Because thermal
143
Numerical analysis of deformation and thermal behavior during ultrasonic Al ribbon bonding
144
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Thermodynamic crystal) (Heat and stress flow) (Stereolithography) (CAD/CAM/CAE)
145
Visualization of heat and stress flows in thermodynamic crystals fabricated by laser scanning stereolithography
by using the stereolithography. The spatial resolution in the intermediate layers. The heat and stress flow patterns in the
fabricated body was approximately 0.5 % in size. Figure fabricated thermodynamic crystals were simulated as shown
2(a) shows (100) planes of the sintered body centered cubic in Fig. 3(a) and (b), respectively. The heat flow
structure composed of the micrometer order alumina transmitted effectively from the source plane to the opposite
lattices including the air spheres. Deformation and cracking side along the connected copper spheres of the percolated
were not observed. Linear shrinkage on the horizontal axis metal phases with high thermal conductivities. And, the
was 6.5 % and that on the vertical axis was 7.8 %. It is tensile stresses were distributed periodically into the
possible to obtain the uniform shrinkage by designing an alumina lattices of the distributed ceramics phases with high
appropriate elongated structure in the vertical direction for Young's modulus. Through introductions of point or plane
compensating the gravity effect. The relative density defects in to the perfect periodic structures, the heat and
reached at 98.5 %. Dense alumina microstructure was stress flows can be localized into the specific regions to
formed, and the average grain size was approximately 2 μm. control the thermal and mechanical properties of the
Figure 2-(b) shows the (100) plane in the sintered artificial crystals.
thermodynamic crystal composed of the pure copper
spheres and alumina matrix. Diffraction peaks indicating 4. Conclusions
the formation of copper oxide phases did not appear in X- Three dimensional thermodynamic crystals with body
ray diffraction spectra. Sintering defects of cracks and pores centered cubic structures composed of pure copper spheres
were not observed in the microstructures of the copper arrangements in alumina matrices were fabricated by using
phase. Smooth interfaces between the copper spheres and stereolithography and heat treatment processes. The
alumina matrices were obtained successfully. These periodic arrangements with high part accuracies of dense
material phases were joined continuously without
Fig. 2 A sintered alumina lattice with the inverse body centered cubic structure fabricated by using stereolithography of a
structural joining process (a) and a formed thermodynamic crystal with periodic arrangements of pure copper spheres in an
alumina matrix though powder sintering treatments (b).
Fig. 3 The computer graphic models of simulated heat and stress distributions in the fabricated thermodynamic crystals
with the body centered cubic structure. The graphics (a) and (b) show the heat and stress flows through the connected metal
spheres and the continuous ceramic matrix, respectively.
146
Transactions of JWRI, Vol.39 (2010), No. 2
alumina and copper phases were observed and verified by for Advanced Structural and Functional Materials Design)
using a digital optical and scanning electron microscope. from the Ministry of Education, Culture, Sports, Science
The heat and stress distributions in the thermodynamic and Technology (MEXT), Japan.
crystals were simulated and discussed theoretically by finite
difference time domain methods. The heat flow transmitted References
effectively through the connected copper spheres with high [1] C. Kai and L. Fai: Rapid Prototyping, Singapore John Wiley
thermal conductivities, and the tensile stress were and Sons, (1997).
distributed in the alumina matrices periodically with high
[2] A. Pique, A.Holmes and D. Dimos: Materials Research
Young's modulus. Society Symposium Proceedings, (2003), pp.758-764.
5. Acknowledgments [3] S. Kanehira, S. Kirihara and Y. Miyamoto, Journal of the
American Ceramic Society, 88 (2005), pp.461-1464.
This study was supported by Priority Assistance for the
Formation of Worldwide Renowned Centers of Research - [4] S. Kirihara and Y. Miyamoto, The International Journal of
The Global COE Program (Project: Center of Excellence Applied Ceramic Technology, 6 (2009), pp.41-44.
147
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Visualization) (Microwave) (Photonic crystal) (Stereolithography) (Powder sintering process)
2. Experimental Procedure
The photonic crystal structure was designed as an Fig. 2 A schematic illustration of stereolithography
inverse type with the periodic arrangements of air spheres system of a computer aided design and manufacturing.
in a pure copper bulk by using a computer graphic
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* Graduate school, Osaka University, Suita, Japan Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
** JWRI, Osaka University, Suita, Japan Japan
148
Visualizations of microwave emissions through pure copper photonic crystals
4. Conclusion
We succeeded in fabrications of pure copper photonic
crystals with or without compressed lattice spacing by
using stereolithography and a powder sintering process.
Fine metal structures were precisely formed without
delamination and fracture. The obtained photonic crystals
showed perfect band gaps from 95 to 110 GHz in frequency.
The modified crystal with compressed lattice spacing
Fig. 3 Pure copper photonic crystals with the inverse
exhibited directional transmission properties at the
body centered cubic structure fabricated by using the
frequency range from 105 to 110 GHz. Simulated and
stereo-lithography and powder sintering processes.
measured transmission properties of the microwave
propagation through the copper crystals showed good
149
Transactions of JWRI, Vol.39 (2010), No. 2
agreements. The compressed metal crystals are considered for Advanced Structural and Functional Materials Design)
to be capable of controlling the emission of high energy from the Ministry of Education, Culture, Sports, Science
microwaves in the near future industrial fields. and Technology (MEXT), Japan.
References
Acknowledgments
[1] E. Yablonovitch: J. Opt. Soc. Am. B, 10, (1993), pp.283.
This study was supported by Priority Assistance for the [2] M. Notomi, A. Shinya, S. Mitsugi, E. Kuramochi and H-Y.
Formation of Worldwide Renowned Centers of Research - Ryu: Optics Express,12 , (2004), pp. 1551-1561
The Global COE Program (Project: Center of Excellence [3] F. Miyamaru and M Hangyo, appl. Phys., 84,(2004), pp. 2742,
150
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Photonic crystal) (Terahertz waves) (Micro stereolithography) (Transmission line modeling)
(Localized mode)
Fig. 1 Schematic illustrations of a unit cell of diamond structure (a), components of resonance cells (b), and a terahertz wave
resonator (c). The resonator is composed of two diamond photonic crystals to localize the terahertz waves in an aqueous phase.
151
Visualizations of terahertz frequency amplifications in water cells introduced into alumina diamond photonic
crystals
152
Transactions of JWRI, Vol.39 (2010), No. 2
153
Transactions of JWRI, Vol.39 (2010), No. 2
:HOGSURFHVVPDQDJHPHQWXVLQJGDWDYLVXDOL]DWLRQ
$SSURDFKHVWRFDOFXODWLRQDQGGLVSOD\RIZHOGLQJVWDWLVWLFDOLQIRUPDWLRQ
PXVWEHVWUDLJKWIRUZDUGWRXVHDQGPDLQWDLQLQDSURGXFWLRQ :6 ¦ V X (1)
HQYLURQPHQWZKHUHZHOGLQJSURFHGXUHVDUHRIWHQFKDQJHG X
7KLV ZRUN IRFXVHV RQ VRPH NH\ LVVXHV ZLWKLQ WKLV The factor of 10-5 is included for numerical convenience.
HQRUPRXV WDVN 6LJQDWXUH LPDJH GDWD REMHFWV >@ KDYH To investigate the behavior of :S along a weld,
SURYHQ WR EH XVHIXO IRU UHDOWLPH IDXOW GHWHFWLRQ ZLWK signatures were added successively to a signature set, and
XQDVVLVWHGHOHFWULFDOVHQVLQJDQGDXWRPDWHGGHWHFWLRQZLWK :S was calculated at each stage. Figure 1 plots :S for the
PLQLPDO RSHUDWRU LQWHUYHQWLRQ LV RQ WKH KRUL]RQ >@ +HUH control weld and three of the faults.
WKH VWDWLVWLFDO SURSHUWLHV RI WKH VLJQDWXUHV WKHPVHOYHV DUH
LQYHVWLJDWHG IRU DSSOLFDWLRQ WR ZHOGLQJ PDQDJHPHQW LQ
VRPH VLPSOH FDVHV 7R GHYHORS DQG HYDOXDWH D
FRPSUHKHQVLYH LQGXVWU\ UHOHYDQW GDWD DQDO\VLV V\VWHP LW LV
LPSRUWDQWWRKDYH DODERUDWRU\IDFLOLW\DYDLODEOHZKLFKFDQ
HPXODWH WKH DXWRPDWHG ZHOGLQJ UDWH RI LQGXVWU\ EXW ZLWK
IXOO FRQWURO RI WKH ZHOGLQJ HQYLURQPHQW 7KLV ZRUN DOVR
GHVFULEHVRXUSURJUHVVLQEXLOGLQJDIDFLOLW\IRUWKLVSXUSRVH
([SHULPHQWV
:HOGVZHUHEHDGRQSODWHZLWKWZRPPORQJPP
GHHSPLOGVWHHOFRXSRQVSRVLWLRQHGDORQJWKHZHOG)LJ
ZKHUHWKHZHOGVEHFDPHRYHUODSMRLQWV 3XOVHZHOGLQJZDV
HPSOR\HGZLWKDFRQWDFWWLSWRZRUNGLVWDQFHRIPPWKH
ZHOGLQJKHDGZDVVHWDWDQDQJOHRIRWRWKHKRUL]RQWDOLQ
WKHWUDQVYHUVHSODQHPPZLUHZDVXVHGZLWKZLUHIHHG
VSHHG PPV WKH WUDYHO VSHHG ZDV PPV DQG WKH
VKLHOGLQJJDVZDVDUJRQEDVHGZLWKDSSUR[LPDWHO\&2
DQGR[\JHQDWDIORZUDWHRI/PLQ)RXUIDXOWVZHUH
FRQVLGHUHGZLUHMDPEHIRUHWKHILUVWFRXSRQFRQWDPLQDWLRQ
WKHSDUWVZHUHQRWFOHDQHGSULRUWRZHOGLQJPLVSODFHGSDUWV
D VLQJOH FRXSRQ LQ WKH PLGGOH RI WKH ZHOG DQG QR SDUWV Fig. 1 Cumulative signature image scatter for several faults.
154
:HOGSURFHVVPDQDJHPHQWXVLQJGDWDYLVXDOL]DWLRQ
6WDWLVWLFDOFDOFXODWLRQV
For the control weld (Fig. 1a), :S increases during
3GLUHFWLRQ
incorporated in the signature image set. For the weld with a
FRXSRQVLJQDWXUHV
single coupon (Fig. 1b) there is an increase at the coupon as FRXSRQVLJQDWXUHV
expected, while the wire feed jam gives a very large scatter FRXSRQSURMHFWLRQ
FRXSRQSURMHFWLRQ
starting at the time of the problem and decreasing as more
155
Transactions of JWRI, Vol.39 (2010), No. 2
&RQFOXVLRQV
The use of statistical measures with signature images,
leads to the possibility of sophisticated analysis treating
both inter- and intra-weld variations in complex robotic
weld sequences. Using more sophisticated measures than
(:S)total will allow detection of a range of faults as well as
indicators of reproducibility and drift. :KHQFRPELQHGZLWK
DQ DXWRQRPRXV GDWDEDVH SURFHVVLQJ JLJDE\WHV RI ZHOGLQJ
GDWD LQ WKH EDFNJURXQG RQ D PXOWLFRUH FRPSXWHU WKHVH
PHDVXUHV FDQ SURYLGH WKH EDVLV IRU D SUDFWLFDO ZHOG
PDQDJHPHQW V\VWHP &RPSXWLQJ KDUGZDUH ZLWK WKH
QHFHVVDU\ FDSDELOLWLHV LV UHDGLO\ DYDLODEOH 7KH DUHD ZKHUH
VXEVWDQWLDOUHVHDUFKLVUHTXLUHGLVLQGHULYLQJWKHQHFHVVDU\
)LJ 'UDZLQJ VKRZLQJ ODVHU FXW SODWH ZLWK SQHXPDWLF PDWKHPDWLFV DQG DOJRULWKPV FRXSOHG WR DQ DSSURSULDWH
FODPSLQJ DUUDQJHPHQW 7KH ZHOGLQJ KHDG WUDYHOV YLVXDOL]DWLRQLQWHUIDFHIRUZHOGLQJPDQDJHUV
EHWZHHQ WKH EUDVV FODPSLQJ EDUV HDFK RI ZKLFK
SURYLGHV N1 FODPSLQJ ORDG 3LQV LQ WKH EDVH SODWH 5HIHUHQFHV
UDLVHWKHWDJVWRSURYLGHVKRXOGHUVIRURYHUODSZHOGV >@6: 6LPSVRQ 6FL 7HFKQRO :HOG DQG -RLQLQJ
SS
>@6: 6LPSVRQ $XVWUDODVLDQ :HOG - ± :HOG 5HV 6XSS
SS
156
Transactions of JWRI, Vol.39 (2010), No. 2
1. Introduction longitudinal [2], Rayleigh [3], and Lamb [4,5] waves. Each
Recent experimental studies and new physical models of these has its own advantages and unique applications. In
[1] demonstrate the high potential of nonlinear ultrasonics particular, the nonlinear Lamb waves are certainly the best
to quantitatively characterize fatigue and creep damage in candidate for efficiently performing a long range inspection
metals. On the fundamental side, a physics-based model is of plate and shell structures. Due to the dispersive and
needed to establish the relationship between the measured multi-modal characteristics of the Lamb waves, the
higher order harmonic and the parameters that characterizes measurement of the second harmonic wave is more difficult
fatigue damage. Recent advances in metal fatigue research than in the non-dispersive waves. A theoretical analysis on
have enables predicting, with reasonable accuracy, the the conditions to generate a spatially growing higher
cumulative plastic deformation in a structure component harmonic has been performed [6]. In addition, an advanced
during cyclic loading. Thus, a physics-based relationship signal processing technique has been applied to extract the
between the higher order harmonic and the cumulative fundamental and second-harmonic amplitudes of a
plastic strain would allow us to estimate the remaining measured Lamb wave signal.
fatigue life of a component based on the nondestructive The developed techniques have been applied to
nonlinear ultrasonic measurements. characterize damage in materials subject to static load, low
This paper presents our recent experimental and theoretical and high cycle fatigue for various metallic materials.
developments in the nonlinear ultrasonic material For example, Fig. 1 shows results of measured material
characterization technique. The developed experimental nonlinearity for the nickel-base superalloy (IN-100) during
techniques are applied to track fatigue and other plasticity- a low cycle fatigue using both longitudinal and Rayleigh
induced damage in nickel-base and aluminum alloys to waves. The maximum stress level is 105% of the yield
demonstrate the capability of the techniques. The stress, and the fatigue tests are interrupted to perform the
microplasticity-based model that relates the acoustic nonlinear ultrasonic measurements at different numbers of
nonlinearity and the cumulative plastic strain is introduced fatigue cycles. Figure 1 shows a rapid increase in İ ŰġĩŶűġ
and results are compared with experimental data. to 30%) during the first 40% of fatigue life, which
demonstrates that these nonlinear ultrasonic measurements
2. Experimental results can be used to quantitatively characterize the damage state
We have developed robust experimental techniques to of this material in the early stages of fatigue life. Similar
track the evolution of fatigue damage in several different trend of large increases in the initial stage is observed in the
alloys including the nickel-base superalloy (IN-100) and Lamb wave measurement results. This implies that the
aluminum alloys (AA6061-T6, AA2024-T3, Al1100-H14), dispersive nature of Lamb waves does not alter their
with the acoustic nonlinearity parameter and demonstrated interaction with fatigue damage and that there is a
its effectiveness and robustness by making repeatable fundamental relationship between fatigue damage and
measurements of the material nonlinearity parameter ( ĪġŪůġ acoustic nonlinearity, independent of wave type. Figure 2
multiple specimens subjected to static, high- and low-cycle shows the acoustic nonlinearity (versus the plastic strain)
fatigue load. Varying geometries of the structure measured with Lamb waves for Al-1100-H41 specimens
components to be inspected requires the experiment method subjected to low cycle fatigue. Note that Fig. 2 can be
to adapt to the geometrical constraints. This led to important baseline data for lifetime prediction of a
developments of three different techniques that use component subjected to cyclic fatigue loading.
157
Nonlinear ultrasound and its applications in quality inspection and damage assessment in metallic materials
Longitudinal wave measurement (Specimen #1) The effects of the microplastcity enter in the coefficient
158
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Fatigue damage parameter) (Thermography) (Sacrificial test piece) (Bridge maintenance)
(Crack growth)(Thin steel plate)
1. Introduction
Unit: mm
͆The sacrificial test piece” is used as a specimen Jig-plate
Sacrificial Test Piece
attached to the member of a main structure in order to 㻝㻜 㻜㻚㻡
evaluate the damage before the appearance of a crack in a
member of that structure. The sacrificial test piece is 㻝㻞 㻝㻜
㻠㻜 㻠㻜
designed so that it is damaged earlier than the main 㻡㻜 㻝㻟㻡 㻣㻜 㻝㻟㻡 㻡㻜
members under the same loads because of crack and stress
magnification. The damage to the bridge members can be 㻞㻡
estimated by the observation of the sacrificial test piece. If
the fatigue damage parameter can be made clear by the 㻝㻜㻜 㻞㻜
behavior of the sacrificial test piece, the maintenance
management of the structure can be determined.
Thin steel plates, which have initial cracks at the 㻠㻠㻜
center, are used as the sacrificial test pieces in this study. Fig. 1 Sacrificial test pieces with jig-plate
When strains are applied to the main member, these are
transmitted from the main member to the thin steel plate
and the crack in the thin steel plate will grow as a result.
Therefore, the monitoring of fatigue damage parameters on
the bridge can be carried out by the observation of the
crack growth in the thin steel plate. If the thin steel plate
can be used as the sacrificial test piece, it seems that fatigue
damage on a bridge can be monitored widely. Because the
thin steel plate is cheap, everyone can obtain it easily [1].
In this paper, a measuring method for the crack length
of the sacrificial test piece from a long distance by self-
reference lock-in thermography is proposed. Applicable
stress ranges and frequency of measurement by the
proposed method are clarified.
Fig. 2 Measurement of crack length in sacrificial test
2. Remote Measurement of Crack Length piece by self-reference lock-in thermography.
by Self-Reference Lock-in Thermography
The sacrificial test piece was attached to four steel jig-
plates by bolts. The shape and the dimensions of the jig- be carried out in a short period. The specimen is attached
plates are shown in Fig. 1. The thickness of the sacrificial on the lower flange of bridge members with high strength
test piece is 0.5 mm, and the thickness of one side edge of vices at the edge of the jig-plates.
the jig-plate is 12mm and other part of the jig-plate is If the crack length of the sacrificial test piece can be
10mm. Using the jig-plates, the strain between the measured from a long distance, the damage in a member of
connected points is concentrated at the sacrificial test piece a bridge can be evaluated more easily and cheaply. We
by the difference in stiffness between the thin plate and the propose a measuring method of the crack length from a
jig-plate. Strain in the sacrificial test piece is concentrated long distance by thermoelastic stress analysis.
more than about 3 times that of the flange by theoretical A self-reference lock-in infrared thermography
calculation. This strain concentration makes the crack technique is newly proposed, in which a reference signal is
growth faster, and the measurement in bridge members can constructed by using the same sequential data on
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* JWRI, Osaka University, Osaka, Japan Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
** Graduate School, Kobe University, Kobe, Japan Japan
159
Measurement of fatigue damage parameter by sacrificial test piece and thermography
Crack Tip
Crack Ti
Fig. 3 Crack tip in sacrificial test piece searched Fig. 4 Crack tip in sacrificial test piece searched
by self-reference lock-in thermography by self-reference lock-in thermography
ǻV=120MPa, 10Hz, Crack length: 20.5mm) ǻV=120MPa, 10Hz, Crack length: 38.0mm)
thermoelastic temperature change. Temperature change in a frequency during measurement were 120MPa-10Hz and
region of interest, such as crack tip, is correlated with that 60MPa-3Hz.
in a remote area, where uniform stress is applied, for Figure 3 and 4 show contours drawings of the relative
reference signal construction. The temperature changes thermoelastic temperature change distribution obtain by the
obtained from the region of interest and remote area are in- self-reference lock-in thermography in the measurement of
phase and have similar waveforms but big differences are 20.5 mm and 38.0 mm. The applicable crack length of the
found in their amplitudes. Consequently if a reference sacrificial test piece is from 20 mm to 40 mm [1]. So Fig. 3
signal could be constructed from the signal obtained from a and 4 show the results of just about the lower and upper
remote area, it is possible to perform correlation processing limits of applicable range. In spite of the crack length, the
without an external reference signal. The lock-in algorithm location of the crack tip can be estimated as the largest
based on the least squares method is employed for signal point of the thermoelastic temperature change.
processing under random loading. It enables us to measure Figure 5 shows the comparison between the crack
the distribution of relative intensity of applied stress under lengths measured by crack gauge and those estimated by
random loading without using any external loading signal the self-reference lock-in thermography. Good agreement
[2]. can be found between the crack length measured by crack
To examine the applicability of self-reference lock-in
Crack Length mesured by Thermography (mm)
crack growth measured by crack gauge. Distance from the Fig. 5 Comparison between crack lengths
lens to thin steel plate was 2 m. Loading stress range and measured by crack gauge and thermography
160
Transactions of JWRI, Vol.39 (2010), No. 2
gauge and those estimated by the self-reference lock-in sacrificial test pieces in the bridge members from long
thermography in all ranges, not only in the case of 120MPa- distance by the self-reference lock-in thermography is
10Hz, but also in the case of 60MPa-3Hz. So it can be said demonstrated.
that the proposed method of self-reference lock-in
thermography can measure the crack length of the Acknowledgments
sacrificial test piece accurately. The authors would like to thank Mr. Nishimura and
Mr. Nakatsuji for their help during the experiments.
3. Conclusions
(1) Even when the frequency is as small as 3Hz, the self- References
reference lock-in thermography can measure the crack [1] Y. Sakino, Y.-C. Kim, and Horikawa, Trans. of JWRI 35(1),
length of the sacrificial test piece accurately in the case 63-70.
[2] T. Sakagami, K. Nishimura and S Kubo, Proc. of SPIE
that the stress range is 60MPa or more.
Vol.5782, 379-387.
(2) The possibility of measurement of crack length in the
161
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Thermal spraying) (Ambient pressure) (Disk-shaped splat) (Splash splat) (Adhesion strength)
162
Role of ambient pressure on splat formation and coating adhesion strength during thermal spraying process
163
Transactions of JWRI, Vol.39 (2010), No. 2
1. Introduction
As the basic industry of national economy, power Table 1 ESD Parameters
industry has been the focus of national development. Recent
Frequency
Discharge
Capacity
Weight
two decades, it is the fastest development period and gains
Output
Model
Input
Size
the most achievements [1] [2].
If the generator rotor shaft wear or damage occurs, rotor
running at high speed is hindered and generator output
power decreased, which may result in the units failure [3]. DZ- AC 1.4 40u 40 420 50-
This paper discusses that the repair of worn parts for 1400 220 KVA 60u Kg F 1200
generator rotor shaft at site using DZ 1400 Electro Spark 50 80 HZ
Deposit (ESD) system manufactured by Surface HZ
Engineering Technology Institute of CAAMS with
satisfactory results and successful experiences.
with ground line of ESD, and repair welding on the sample
surface with a thickness of 1mm. After overlaying by ESD,
2. Electric Spark Deposition (ESD)
obtain cross-section of the coating by line cutting method
Electric Spark Deposition Principle
and prepare metallographic sample.
Electric Spark Deposition (ESD) process is to instantly
release in high frequency high-energy from the power Table 2 Chemical composition of the electrodes for sample
supply between electrode and substrate, then produce air
Ni Cr Fe Mo Mn Si Other
ionic passage, inducing micro-zone of high temperature,
high-pressure physic-chemical metallurgical process 70 14 4.5 --- 7.8 0.5 Bar
between high alloy electrode and substrate surface.
Test Analysis
ESD process combines the characteristics of welding,
spraying and other processes with low heat input, good Figure 1 is the cross-section micrograph (u3000) of the
coating. It’s known that:
metallurgical bonding between weld pass and substrate.
1) There isn’t welding defect such as pore, oxide slag
inclusion, crack, etc. in overlay.
ESD Equipment 2) Crystallite of overlay and substrate transition layer is fine,
Power output of ESD (DZ-1400) is 1400W㸪using argon no growing tendency.
for welding, self-consumable anode moves at 4700rps at the 3) The coating structure is very fine small columnar crystal,
worn place of the work piece, induce HF spark discharge, which shows good corrosion and abrasion resistance.
and form dense and homogeneous overlay coating with a Figure 2 is the energy spectrum analysis of Ni, Cr
thickness up to 2mm or more. Substrate temperature elements in the coating. The thickness of heat-affected zone
remains around 60Υduring repair, and the substrate does of ESD is 10ȣm. It demonstrates that a good metallurgical
not distort and undercut. Table 1 is ESD parameters. bonding can be obtained by ESD, also shows that deposit
heat-affected zone is small and welding residual stress can
3. Experiment and Analyses be ignored. The average hardness of coating and heat-
Sample preparation affected zone is very close to the matrix hardness HV220.
According to the common material of rotor shaft,
35CrMoA is selected as the sample substrate material with Result
the size of 25mmu10mmu50mm.Nickel-based high alloy The result of experiments show the metallurgical bond
rod is chosen as electrode material with a specification of between coating and substrate has not welding defect by
I3.2mmu6.0mm, the chemical composition is listed in depositing inconel alloy on substrate 35CrMoA using ESD
Table 2. After pretreatment, connect tightly substrate process. After welding, heat-affected zone is very narrow
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* Chinese Academy of Agricultural Mechanization Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
Science(CAAMS), China Japan
164
Repair heavy-duty generator rotor shaft by electro spark deposition process
(thickness of coating reaches 1mm, but heat-affected zone Visual check overlay after each layer DERXW ȝP
is only 0.01mm), hardness of coating conform to that of thickness) to see if has welding defects such as oxide㸪if
substrate. The above results meet the requirement of rotor has, use needle file to remove and use brass wire brush to
shaft repair in power plant. clean, then overlay the next layer. When overlay is close to
the desired size, use toolmaker's straight edge to align㸪the
overlay should be more than the final repaired size 50-
ȝP$IWHUWKHFRPSOHWLRQRIRYHUOD\LIWKHUHLVZHOGVSRW
or side missed during inspection can make up welding.
Post-treatment
After the completion of ESD, the bench worker will work
on the dimensional recovery and finishing polish of the
repaired surface of the shaft. The repair standard refers to
the technical specifications of shafts. Fig. 3 is seal section
shaft of generator rotor after repair and finishing,
respectively.
165
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Resistance spot welding) (Nugget formation) (Numerical modeling) (Sheet thickness)
† Received on 30 September 2010 *** China Science Lab, General Motors Research &
* State Key Lab of Mechanical System and Vibration, Development Center, Shanghai, China
Shanghai Jiao Tong University, Shanghai, China Transactions of JWRI is published by Joining and Welding
** Shanghai Key Lab of Digital Autobody Engineering, Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
Shanghai Jiao Tong University, Shanghai, China Japan
166
Visualization of nugget formation in resistance spot welding of multi-stackup sheets
nugget sizes were modeled and the results are shown in Fig.
3. The diameter B at the faying interface between top and
middle sheet is always greater than the minimum acceptable
weld size, 5 mm. However, there is a significant decrease in
the diameter A at the faying interface between middle and
bottom sheet. It is shown that, when the thickness of the
bottom sheet is 2.4 mm, the weld size at the faying interface
A is less than the minimum required weld size (i.e.4 mm)
[3].
(a) 240ms
(a) Diameter A
(b) 360ms
(b) Diameter B
Fig. 3 Effect of the bottom sheet gage on nugget diameter (a)
Diameter A, and (b) Diameter B.
4. Conclusions
The conclusions of this study are summarized as
follows:
(1) A finite element model has been developed to predict the weld
(c) 460ms
nugget size for RSW of multiple stacks of steel sheets.
Fig. 2 Temperature field during the welding process for0.6mm (2) The weld nugget on the faying interface of DP600 forms
SAE1004 +1.8mm SAE1004 +1.2mm DP600 under a welding earlier than that on the other interface of SAE1004 in RSW. It
current of 9.5kA. is due to the fact that the greater bulk and contact resistivity of
DP600 would lead to a quicker heat development in the weld
cycle than SAE1004.
167
Transactions of JWRI, Vol.39 (2010), No. 2
Acknowledgement References
This research was supported by the General Motors [1] Harlin N., Jones T.B., Parker J.D.:J. of Materials Processing
Collaborative Research Laboratory at Shanghai Jiao Tong Technology (2003), pp. 448–453.
University, the Research Fund of State Key Lab of MSV, [2] Jie Shen, Yansong Zhang, Xinmin Lai and P.C. Wang,
Modeling of Resistance Spot Welding of Multiple Stacks of
China (Grant No. MSV-2010-04) and Project 50905111
Steel Sheets’, Materials & Design, (In Press)
supported by National Natural Science Foundation of China. [3] ANSI/AWS/SAE/D8.9-97. Recommended practices for test
methods for evaluating the resistance spot welding behavior of
automotive sheet steel materials. AWS, 1997
168
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (solder paste) (wettability) (laser displacement meter) (reflow soldering) (lead-free solder)
(surface mount technology)
1. Introduction (2) The equipment has a system which can take a picture of
Built-in electronic parts are made minute as mobile continuous animation of joints with CCD camera in the
products are miniaturized. Then, surface mount technology reflow process.
by reflow soldering is growing rapidly [1, 2]. Moreover, it (3) The equipment has a system which controls the
is known that the wettability to the electrode of the solder temperature with the halogen heater and a system which
paste is affected worsened by the world trend of Pb free in controls the cooling speed by introducing atmosphere
the solder paste [3, 4]. Therefore, the development of the gas into the chamber.
evaluation technique of the wettability of the solder paste is
a pressing need. However, the evaluation technique of the 3. Results and discussion
wettability of the solder paste has only a wetting balance Figure 2 shows the heat pattern in reflow process used.
method in which the size effect of the electronic parts can The shape of the metal mask and copper land is as shown in
hardly be considered [5-7]. So, the development of Fig. 3. The 1608 chip capacitor was used for electronic
equipment which can evaluate the wettability of the solder parts. The results of measuring the descent behavior of
paste to actual electronic parts at the speed of parts in the electronic parts are shown in Figs. 4-7.
reflow process was attempted with a laser displacement In the case of Sn-3.0Ag-0.5Cu (mass%) solder paste, it
meter in this research. has been understood that electronic parts descend with the
melt of the solder powder as shown in Figs. 4 and 5. In
2. Feature of developed equipment addition, the descent of electronic parts has progressed
As shown in Fig. 1, the equipment developed by this when the molten solder was wetting up to the electrode on
research is chiefly composed of the following three parts. sides of electronic parts. Even if Sn-3.5Ag-8In-0.5Bi
(1) The equipment has the temperature measurement system (mass%) solder paste was used, a similar tendency was
with a thermo-couple and the height measurement observed as shown in Figs. 6 and 7. However, it was
system of electronic parts with a laser displacement understood that Sn-3.0Ag-0.5Cu solder was longer
meter. compared with Sn-3.5Ag-8In-0.5Bi solder for the time to
get to the bottom after electronic parts descended rapidly. It
is thought that this factor depends on the amount of solder
which is wetting up to the electrode on sides of electronic
parts (see Figs. 5 and 7).
Fig. 1 Outline chart of developed equipment. Fig. 2 Diagram of heat pattern in reflow process.
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* Graduate School, Gunma University, Kiryu, Japan Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
** Sanyo Seiko Manufacturing Co., Ltd., Hachioji, Japan Japan
*** Yamanashi Pref. Industrial Technol. Center, Kofu, Japan
169
Development of wettability evaluation equipment for solder paste using laser displacement method
170
Transactions of JWRI, Vol.39 (2010), No. 2
(1) The wettability of the solder paste is appreciated by [3] E. E. M. Noor, N. M. Sharif, C. K. Yew, T. Ariga, A B. Ismail
measurement of the descent speed of electronic parts in and Z. Hussain: J. Alloys and Compounds, 507 (2010),
the reflow process with a laser displacement meter. pp.290-296.
(2) The developed equipment can evaluate the difference of [4] D. Rocak, S. Macek, J. Sitek, M. Hrovat, K. Bukat and Z.
Drozd: Microelectronics Reliab., 47 (2007), pp.986-995.
the wettability of the solder paste which depends on the
[5] J. Kloeser, P. Coskina, R. Aschenbrenner and H. Reichl:
solder paste type. Microelectronics Reliab., 42 (2002), pp.391-398.
[6] K. M. Martorano, M.A. Martorano and S.D. Brandi: J. Mater.
Proces. Technol., 209 (2009), pp.3089-3095.
References
[7] E. Attar and C. Korner: J. Colloid and Inter. Sci., 335 (2009),
[1] B. Noh, J. Choi, J. Yoon and S. Jung: J. Alloys and
pp.84-93.
Compounds, 499 (2010), pp.154-159.
[2] T. Chang, Y. Hsu, M. Hon and M. Wang: J. Alloys and
Compounds, 360 (2003), pp.217-224.
171
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (solder paste) (wettability) (laser displacement meter) (reflow soldering) (lead-free solder)
(modification) (organic acid)
1. Introduction
In surface mount technology, the deterioration of the
wettability to the Cu pad and the electrode of the part of the
solder paste according to the abolition of the use of Pb have
been pointed out [1, 2]. The deterioration of the wettability
influences the reliability of products with solder joints.
Therefore, the improvement of the wettability is a very
important technology in the electronic industry. In this
research, the verification of the effect to the wettability was
attempted by using the surface modification method with an
organic acid from which the effect of decreasing the solid-
state bonding temperature of Sn and Cu was confirmed [3]. Fig. 2 Diagram of heat pattern in reflow process.
The organic acids used in the experiment are formic acid
process used the heat pattern shown in Fig. 2. The
and citric acid. For the comparison, the surface which had
wettability was evaluated by measuring a descent speed of
been processed by the procedure of making to cleanness on
the fragment in reflow process with a laser displacement
the Cu surface used for wettability evaluation method of the
meter.
solder paste (JIS Z3284) was examined.
172
Examination of improvement effect of surface modification of Cu with organic acid on solder paste
wettability using a laser displacement meter
Fig. 5 Relation between descent behavior and reflow Table 1 Effect of surface modification on descent
temperature (modificated by formic acid). behavior.
173
Transactions of JWRI, Vol.39 (2010), No. 2
4. Conclusions References
The conclusions of this study are summarized as [1] D. Rocak, S. Macek, J. Sitek, M. Hrovat, K. Bukat and Z.
follows. Drozd: Microelectronics Reliab., 47 (2007), pp.986-995.
(1) The difference of the wettability of the solder paste [2] J. Kloeser, P. Coskina, R. Aschenbrenner and H. Reichl:
Microelectronics Reliab., 42 (2002), pp.391-398.
which depends on the surface modification of Cu is
[3] S. Koyama and I. Oya: J. Japan Inst. Metals, 73
revealed by the measurement of the descent speed of a
(2009), pp.809-815.
Cu fragment with a laser displacement meter. [4] S. Poulston, R. P. Holroyd, M. Bowker, S. F. Parker and P. C.
(2) The wettability of the solder paste to the Cu surface is H. Mitchell: Surface Science, 402-404 (1998), pp.599-603.
rapidly improved by a surface modification process [5] M. A. Mohamed, A. K. Galwey and S. A. Halawy:
which uses an organic acid. Thermochimica Acta, 411 (2004), pp.13-20.
174
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (High power fiber laser) (MAG arc) (Hybrid welding) (T-joint fillet weld) (Welding direction)
2. Experimental procedure
(a) Vertical plate Horizontal T-fillet welding of SS400 steel plate with the
Arc dimension of 150mm x 50mm x 12mm in thickness was
cond ucted in T-fillet joint by using 10kW class high
Laser 45㼻 power fiber laser and pulsed MAG arc as shown in Fig. 1
(a) and (b), which shows the schematic illustration of the
setup of a hybrid welding. During welding, the behavior of
Laser angle molten metal flow in the weld pool was observed by using
Vertical offset an infrared high-speed video camera with 2000 frames/s.
After the welding, the welded specimens were cut, polished
1.5mm and etched to reveal the microstructure and weld defect.
Horizontal plate
3. Results and Discussions
(b) In hybrid welding, welding direction which means laser-
Arc leading or arc-leading is one of the important and special
Laser
parameters for hybrid welding. Figure 2 shows the typical
bead appearance and cross-section of 1 pass-1 side fully
Arc-leading 35㼻 Laser-leading
penetrated T-fillet weld beads in each welding direction at
the same welding condition. Comparing two weld beads
large differences were easily observed. In laser-leading, a
smooth weld bead was obtained, while a humping and
Vertical plate 2mm undercut bead with pit and void was observed in the arc-
Horizontal plate leading, which was caused by different types of molten
metal flow. In order to reveal this, the behavior of molten
Fig. 1 Schematics illustration of the setup of a hybrid metal flow in the weld pool was observed. Figures 3 and 4
welding show typical frames of weld pool surface during hybrid
(a) Horizontal view , (b) Vertical view welding in laser-leading and arc-leading, respectively. In
the laser-leading molten metal flew forward to the front end
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* Graduate School, Osaka University, Suita, Japan Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
** JWRI, Osaka University, Ibaraki, Japan Japan
175
Effect of welding direction on weld bead formation in high power fiber laser and MAG arc hybrid welding
Arc torch
Pit Wire
Laser beam
Welding direction
Molten pool
of the weld pool, and then went rearward along the edge of
the weld pool. On the other hand, molten metal flew
continuously and rapidly rearward to the rear end of the
weld pool in the arc-leading and in addition, opening of a
pit was observed. Figure 5 shows schematics of these
obvious differences in molten metal flow comparing in
laser-leading and arc-leading, which is caused by the strong
plasma gas flow of arc and its flow direction related to the Fig. 6 Appearance and cross-section of hybrid weld bead of
inclination of GMA arc torch. In high speed arc welding, it SS400 with different welding speeds at laser-leading
176
Transactions of JWRI, Vol.39 (2010), No. 2
4. Conclusions Acknowledgment
The conclusive remarks are summarized as follows. This study was carried out as a part of research activities
(1) 1 pass-1 side fully-penetrated horizontal T-fillet weld of “Fundamental Studies on Technologies for Steel
bead can be successfully made by optimizing the fiber Materials with Enhanced Strength and Functions” by
laser and pulsed GMA arc hybrid welding parameters. Consortium of JRCM.
(2) Different molten metal flows are observed in laser-
leading and arc-leading, which caused differences in References
appearance and cross-sectional shape of weld bead as [1] M. Kutsuna, Welding Guide Book 6, Development in
well as weld defect formation. Advance Welding Processes, JWS, 2009, I-45 - I-66.
[2] M.H. Shin, K. Nakata, Quarterly Journal of the Japan Welding
(3) Laser-leading in the welding direction is beneficial to
Society, 27, 2009, 80-84(in Japanese).
make a smooth weld bead without weld defect at a [3] Japan Welding Society, Handbook of Welding and Joining,
hybrid welding with horizontal T-fillet joint. Maruzen, 2003, 188-189(in Japanese).
177
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Arc) (Power source characteristic) (Arc length) (Gas tungsten arc) (Numerical analysis)
1. Introduction
Arc properties in Gas Tungsten Arc (GTA) strongly In the numerical models, the tungsten cathode, the arc
depend on welding conditions such as arc current, arc plasma and the water-cooled copper anode are described
length and so on [1-3]. In GTA, the arc current and arc relative to cylindrical coordinates, assuming rotational
voltage are determined by an external characteristic of a symmetry around the arc axis. An arc length varied between
power source and an electrical characteristic of the arc. 2 mm and 7 mm at 10 Hz like Fig. 1. In this model, number
There are two kinds of external characteristics of power of the mesh is constant, however, mesh size changes if arc
sources, namely, Constant Current (CC) and Constant length changes. The arc current is set to be a 150 A for a
Voltage (CV) characteristics. The electrical characteristic CC characteristic. The arc voltage is set to be 11.0 V for a
of the arc depends on the arc length. In this study, CV characteristic. The other approximations, governing
dependences of the arc properties and relationships between equations and boundary conditions are given in detail in our
the current and the voltage on the arc length in GTA previous papers [3-5].
employing power sources with the CC and the CV
characteristics were analyzed. In the experiment and the 4. Results and Discussion
numerical simulation, arc length varies time-dependently. Figure 2 shows dependence of arc voltage on arc length
for the CC characteristic. The experimental result of arc
2. Experimental method voltage is between 11.8 V and 15.0 V and the calculation
In the experiments, the free-burning arcs were generated result is between 8.3 V and 12.7 V. Although the
between a tungsten cathode and a water-cooled copper experimental result of arc voltage is higher than the
anode in argon at atmospheric pressure. A commercial calculation result, both results decrease with decrease of arc
GTA torch was fixed on vibratory equipment, and an arc length. Figure 3 shows dependence of arc current on arc
length varied between 2 mm and 7 mm at 10 Hz. An arc length for the CV characteristic. The experimental result of
current was set to be DC 150 A with a CC characteristic. arc current is between 114 A and 189 V and the calculation
An arc voltage was set to be 13.8 V in order to make an arc result is between 80 A and 351 A. Although there are some
current about 150 A with a CV characteristic. difference between the experimental result and the
calculation result of arc voltage, both results increase with
3. Simulation model decrease of arc length and changed circularly. In the same
arc length, arc current at decreasing pattern is higher than
increasing pattern. This is because the thermal problem that
the temperature of the tungsten cathode changes due to
change of arc current.
Figure 4 shows dependence of arc power on arc length.
In experimental results, arc power is between 1776 W and
2248 W for the CC characteristic and is between 1567 W
and 2615 W for the CV characteristic. In calculation results,
arc power is between 1252 W and 1911 W for the CC
characteristic and is between 881 W and 3869 W for the
CV characteristic. With the same variation of the arc length,
the arc power hardly changes for the CC characteristic,
although the arc power largely changes for the CV
characteristic. Furthermore, the arc power decreases with
decrease of the arc length in case of the CC characteristic
and the arc power increases dramatically with decrease of
Fig. 1 Assumption of time-dependence of arc length at the arc length in case of the CV characteristic.
10Hz
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* JWRI, Osaka University, Osaka, Japan Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
Japan
178
Numerical analysis on effects of power source characteristics on arc properties in gas tungsten arc
5. Conclusions
The conclusions of this study are summarized as follows.
(1) Study of dependences of the arc properties and
relationships between the current and the voltage on the
arc length in GTA employing power sources with the
CC and the CV characteristics were experimentally and
numerically analyzed.
(2) It was found that the voltage and power of the arc
decreased with decrease of the arc length in case of the
CC characteristic and the current and the power of the
arc increase dramatically with decrease of the arc length
in case of the CV characteristic.
(3) With variation of the arc length, the arc power hardly
changes for the CC characteristic, although the arc
power largely changes for the CV characteristic.
i) Experimental results
179
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (AC TIG welding) (Cathode spots) (Aluminum) (Oxide layer)
1. Introduction the aluminum plate was recorded with GX1 at frame rate of
Since it is necessary to remove the the oxide layer on 40,000fps for 1s from the arc ignition. On the other hand,
aluminum for realizing high quality weld joint in arc traces of cathode spots were recorded with HPV-1 at frame
welding of aluminum plates, AC TIG welding is generally rate of 500,000fps for ȝV
employed. In duration of EP (Electrode Positive) in AC
TIG welding, the oxide layer is removed by the cleaning
action of cathode spots. In this study, the cleaning action of
cathode spots on the oxide layer in AC TIG welding of the
aluminum was experimentally observed with a high speed
video camera and the cleaning rates of the oxide layer and
velocities of cathode spots were evaluated.
2. Experimental procedure
Fig. 1 shows a schematic illustration of the experimental
setup which consist of a TIG torch, an AC power source
(DAIHEN: DA300P), a water cooled copper electrode, an
aluminum plate as a base metal (A1050, 10mm thickness),
high speed video cameras (SHIMAZU: HPV-1 and NAC:
GX1), a PC for image processing and a data logger
(KEYENCE: NR-200). The tungsten electrode with added
2% La 2 O 3 with the diameter of 3.2mm and the conical
angle of 60 degrees was employed. The electrode gap was
5mm. An arc current was set to be AC 250A (frequency of
70Hz). The length of EP to that of a cycle was 0.3.
Shielding gas was argon and introduced at the flow rate of
15l/min. The wave form of the current was recorded with
the data logger. The cleaning zone of the oxide layer by
180
Experimental observation of cleaning action of cathode spots in AC TIG welding of aluminum plates
181
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Numerical simulation) (Pulsed TIG arc) (Heat source property)
1. Introduction 5DGLDOPRPHQWXPFRQVHUYDWLRQHTXDWLRQ
The heat source properties of TIG arc strongly depend on w Uv r 1 w w wP
(rUv r2 ) ( Uv z v r ) j z BT
wt r wr wz wr
the composition of shielding gas. For example, since the arc
1 w wv w wv wv v
column is constricted due to low electrical conductivity of ( 2 rK r ) (K r K z ) 2K r2
r wr wr wz wz wr r
the helium arc, heat flux onto a base metal in case of helium
TIG arc is higher than that of argon TIG arc. The heat $[LDOPRPHQWXPFRQVHUYDWLRQHTXDWLRQ
source properties can be controlled also by current w Uvz 1 w w wP
( rUvr vz ) ( Uvz2 ) jr BT
waveform. Pulsed TIG welding is suitable for back-bead wt r wr wz wz
welding and thin plate welding, because the heat flux onto w wv 1 w wv wv
(2K z ) (rK r rK z ) Ug
the base metal can be controlled by adjusting peak / base wz wz r wr wz wr
current ratio and frequency. A number of results on (QHUJ\FRQVHUYDWLRQHTXDWLRQ
experimental and theoretical investigations of the heat w Uh 1 w w 1 w rN w h w N wh
(rUvr h) ( Uvz h) ( ) ( )
source properties of DC TIG arc have been reported. wt r wr wz r w r cp w r w z cp w z
However, those of pulsed TIG arc are still not fully jr E r j z E z U
understood because of the complexity of the phenomenon.
&XUUHQWFRQWLQXLW\HTXDWLRQ
In this study, the heat source properties of pulsed TIG arc
1 w
rj r w j z 0
for various shielding gas composition were numerically
analyzed.
r wr wz
2. Simulation Model 2KP̓VODZ
The calculation region for TIG arc consists of a tungsten jr VE r , j z VE z
cathode with diameter of 3.2mm and tip angle of 60 degrees,
arc plasma and a water-cooled copper anode. It is described
in a frame of cylindrical coordinate with axial symmetry
around the arc axis. The electrode gap is set to be 5mm.
Argon (Ar), helium (He) or carbon dioxide (CO 2 ) is
introduced as shielding gas from the upper boundary at the
flow rate of 10 l min-1. For example of physical properties
of each gas, specific heat, thermal conductivity and
electrical conductivity are shown in Fig. 1. Figure 2 shows
current wave form with the peak current of 150A and the
base current of 50A. The frequency is 100Hz and the pulse
width is 5ms. The laminar flow is assumed, and the arc
plasma is considered to be in the local thermodynamic
equilibrium (LTE). The governing equations (1)-(6) are
solved iteratively by the SIMPLEC numerical procedure [1].
The other numerical modeling methods are given in detail
in our previous paper [2].
0DVVFRQWLQXLW\HTXDWLRQ
wU 1 w w
(rUv ) ( Uv ) 0
wt r wr r
wr z
182
Numerical simulation of heat source properties of pulsed tungsten inert gas arc
4. Conclusions
(1) In the case of Ar, the heat caused by Joule heating near
the cathode tip was transported toward the anode by
convective flow and was also transported in a radial
direction by thermal conduction.
(2) In CO 2 , although the heat transport toward the anode
by convective flow was seen, that in the radial direction
due to thermal conduction was smaller than that of Ar
because of influence of large specific heat.
(3) In the case of He, the temperature distribution changed
momentarily following the change in the arc current
due to high thermal conductivity of He.
References
[1] S.V. Patanker: Numerical heat transfer and fluid flow,
Hemishpere Publishing Corporation (1980).
[2] M. Tanaka, et.al.: Plasma Chem. Plasma Process, 23 (2003),
585-606. Fig. 5 Temperature distribution (He)
183
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Arc) (TIG welding) (Numerical simulation) (Unified model) (Emitter material)
Fig. 1 Appearances of W cathode and W-2% ThO 2 cathode Fig. 2 Simulation region
184
Numerical analysis of weld pool formation mechanism in TIG welding in consideration of influence of emitter
material adding to tungsten cathode
given in detail in our previous paper [4]. The area of cathode. Although the maximum flow velocity in the
current attachment on the cathode surface was modeled by outward radial direction close to the anode surface was 11.2
the same treatment in Ref [5]. The cathode surface has the m/s in the 2%ThO 2 -W cathode, that decreased to 5.2 m/s in
work function of the emitter material where the surface a 2%ThO 2 -W cathode. For the reason of the difference in
temperature exceeds the melting point of the emitter weld pool shapes, it is considered that because shear force
material. Except for this region, the surface has the work acting on the weld pool surface for 2%ThO 2 -W cathode
function of the tungsten. was larger than that of a W cathode due to higher cathode
jet velocity, the heat transport in the weld pool near the
surface in radial outward direction increased.
4. Conclusions
It was found that although the maximum cathode jet
velocity in case of a 2%ThO 2- W cathode was 497 m/s, that
in case of pure W cathode was only 156 m/s because of low
current density caused by the flattened surface of the
melting cathode tip. In case of a 2%ThO 2 -W cathode, the
depth and the width of the weld pool became shallow and
wider than those in case of a pure W cathode. For the
reason of the difference in weld pool shapes, it is
Fig. 4 Calculation result of temperature for a 2%ThO 2 -W considered that because shear force acting on the weld pool
electrode surface for a 2%ThO 2 -W cathode was larger than that of a
185
Transactions of JWRI, Vol.39 (2010), No. 2
W cathode due to higher cathode jet velocity, the heat [2] S.V. Patanker: Numerical heat transfer and fluid flow,
transport in the weld pool near the surface in radial outward Hemishpere Publishing Corporation (1980).
direction increased. [3] M. Tanaka, et.al.: Plasma Chem. Plasma Process, 23 (2003),
585-606.
[4] M.J. McNallan and T. Debroy: Metall. Trans. B, 22B (1991),
References
[1] M. Tanaka et al: J. Phys. D: Appl. Phys., 38 (2005), pp. 29-35. pp. 557-560.
[5] M. Tanaka et al: IIW Doc. 212-1091-06.
186
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Finite element method) (Efficiency) (Iterative substructure method) (ANSYS) (Residual stress)
x
economic gain can be achieved [1]. The time and cost
associated with reliable analysis of complex industrial
problems appear to be the main reason contributing to this
practice. Investigators have attempted to develop strategies 3 4
10
0
7
10
6 1
focusing on different dominant factors that influence the
10
8 2 5
efficiency of modeling actual welded structures [2–6].
Noting the fact that the high temperature region which exhibits
strong non-linearity is limited to a very small area compared to
y
the size of the model to be analysed and the remaining part is Fig. 1 The positions at which hole-drilling method was used
mostly linear, an iterative substructure method (ISM) has been
to measure the residual stress.(Unit: mm)
developed to transform welding problem into the combination
of a large linear problem and a small but moving strong
nonlinear region, to improve the speed of computation.[7,8] 3. Numerical Model
In this work, we carried out a comparison of calculation Figure 2 shows the coordinate system and the mesh used
efficiency between an in house finite element (FE) code to predict the quasi-stationary shape of the molten pool.
developed based on the idea of iterative substructure Only half of the specimen is considered as it is symmetrical
method (ISM) and commercial FE software ANSYS. To about the symmetry plane of the weld. It can also be seen
achieve the research purpose, thermal elastic-plastic finite from Fig.2 that a dense mesh is put around the heat source
element analysis of laser welding process of TC4 plate was and the vicinity of top surface and bottom surface, while for
taken as an example and was simulated by ANSYS and an the other regions, a coarse mesh is employed. Eight node
in house developed FE code based on ISM, respectively. In brick element is used in this study and the model contains a
total of 20503 nodes and 16800 elements.
order to make a comparison of residual stress between
experiment and simulation, a hole drilling technique was
used to measure the residual stresses in the weldments.
2. Experiment Procedures
Welding was done with a continuous wave CO 2 laser
which delivers 3KW maximum mean power. The diameter
of the minimum focal spot was 0.4mm. The material used in
this study was TC4 of which the chemical composition is
presented in Table 1. The geometry of the welded plate was
200mm×200mm×2mm. Samples were manually cleaned Fig. 2 Mesh of the plate
187
Efficiency comparison between iterative substructure method and commercial software
Material Properties 200 ANSYS
The thermal and mechanical properties of TC4 alloy
employed in this study are shown in Fig. 3. 150
CPU time / h
1.2 100
E× 105[MPa]
1.0
C [J/gK] 50
0.8 VY× 103[MPa] ISM
0.6
0
U× 102[g/mm3]
0.4 Fig. 4 A comparison of CPU time between ANSYS and
0.2 ISM.
D× 104[1/K]
O>JmmsK@
0.0 1000
400
Analysis Procedure 0
Commercial FE software ANSYS and an in house
developed FE code based on ISM were used to solve the 0 5 10 15 20 25 30
between the analysis results for each case, the same analysis Fig.6 Results of transversal stress
procedure was used for both cases.
The problem was treated as a sequentially coupled The experimental result of longitudinal stress at the weld
analysis. First, a thermal analysis was performed to predict center is about 670MPa. The longitudinal stress at the weld
the temperature history of the model. Subsequently, center predicted by ISM and ANSYS are about 800 MPa
temperature results of the thermal analysis were applied as and 920MPa, respectively. In other words, at the central
body loads in a nonlinear transient thermal–mechanical region of the weld, the results predicted by ISM are closer
analysis. The same mesh used in the thermal analysis was to the experimental result than that predicted by ANSYS. In
used for the thermal–mechanical analysis. A volumetric addition, it is noted that when ANSYS is used the
heat source was employed in the thermal analysis. The maximum longitudinal stress appears at the weld center.
continuous welding process was approximated by the When ISM is used, the maximum longitudinal residual
advancing of a volumetric heat source. During welding, the stress arises in the heat affected zone(HAZ), which is
temperature, and consequently the temperature dependent similar to the experimental results. It can also be seen from
material properties, changed very rapidly. Thus, material Fig.6 that ISM predicted a slightly narrower width of the
properties and the stiffness matrix were updated at every tensile stress zone than the experimental result, while
equilibrium iteration. In addition it is assumed that elements ANSYS predicted a tensile stress zone in which the width is
with a temperature higher than 1500 Υ had little appreciably larger than the experimental result.
mechanical effect on the surrounding elements. Furthermore, Fig. 6 represents the distributions of
residual stress on the top surface in the Y-direction, as
4. Results and Discussion measured by hole-drilling method and the finite element
Figure 4 shows a comparison of CPU time between method. As Fig. 7 shows, ISM provided a slightly lower
ANSYS and ISM. As shown in the figure, the computation stress than the experimental results at the centerline of the
speed of ISM is about 13 times faster that that of ANSYS. weld, but are close to the experimental results in the HAZ.
Figure 5 depicts the distributions of longitudinal
residual stress on the top surface as measured by the hole- 5. Conclusions
d r illing metho d and the finite element method. (1) The computation speed of ISM is about 13 times faster
than that of ANSYS;
188
Transactions of JWRI, Vol.39 (2010), No. 2
(2) At the central region of weld, the results predicted by [4] L.E. Lindgren, H.A. Haggblad, J.M.J. McDill, A.S. Oddy,
ISM are more accurate than those predicted by ANSYS; Comput. Methods Appl. Mech. Eng. 147 (1997) 401–409.
(3) The tension stress zone predicted by ISM is narrower [5] Y. Ueda, M.G. Yuan, ASME J. Eng. Mater. Technol. 115
than the experimental result, while that predicted by (1994) (1993) 417–423.
ANSYS is wider than the experimental result. [6] A. Bachorski, M.J. Painter, A.J. Smailes, M.A. Wahab, J.
Mater. Process. Technol. 92/93 (1999) 405–409.
References [7] H. Nishikawa, H. Serizawa and H. Murakawa: Sci. Technol.
[1] L.E. Lindgren, J. Thermal Stresses 24 (2001) 305–334. Weld. Join., 2007, 12, (2), 147–152.
[2] J.M.J. McDill, A.S. Oddy, Comput. Struct. 54 (1995) 183– [8] H. Serizawa, S. Itoh, T. Tsuda, H. Nishikawa, M. Shibahara
189. and H. Murakawa: Proc. Int. Conf. on ‘Welding science and
engineering’, Xi’an, China, October 2005, Xi’an Jiaotong
[3] J.M.J. McDill, J.A. Goldak, A.S. Oddy, M.J. Bibby, Comm. University, 240–246.
Appl. Num. Methods 3 (1987) 155–163.
189
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Weld pool) (Numerical simulation) (Welding distortion) (Penetration depth) (Thermal
conductivity) (GMA welding) (Temperature distribution of base metal) (Robotic welding)
(CAD)
1. Introduction
Recently, the robotic welding is applied in the need field Welding direction
to save the human costs and to reduce the need for skilled +
㼀㼛㼞㼏㼔
workers. This is important to save the production costs. If
the welding distortion before the welding can be predicted,
the design cost will be saved.
The simulation technologies are developed according to
the development of computers. In many applications, the
numerical simulation is applied to save costs. In the field of
the welding, the numerical simulations were proposed and
- 㻿㼑㼚㼟㼛㼞
35
distortion. Therefore, the control of the welding distortion
a
is demanded in the steel structure welding to improve the
200
productivity. For this purpose, the estimation of the amount
160
Bead
35
20
190
Experiment and numerical simulation in temperature distribution and welding distortion in GMA welding
conditions. in Fig. 4(b). The left side of the base metal without the
clamping was bended.
3. Numerical model of welding
Dimension of the numerical model is the same as the
fundamental experiment. The flow of the weld pool was
neglected. The arc was approximated by the heat input. The
numerical model is induced with the assumption, that the
heat conduction is dominant. The heat input is equal to the
product of the welding voltage V and the welding current I.
The calculation was carried out under the assumption that
40% of the heat input was used for melting of the electrode
wire, i.e., 60% of VI was given to the base metal. Let heat Fig. 3 Heat input to metal surface and weld pool
efficiency to the base metal, K be 0.7. The heat input VI is
8400 J/s. An equation of heat conduction is
wU § w 2U w 2U w 2U ·
Uc N ¨¨ 2 2 2 ¸¸ (1)
wt © wx wy wz ¹
where NU and c are heat conductivity, density and specific
heat, respectively.
The distortion analysis is based on the following partial (a) Temperature distribution. (b) Displacement due to
distortion.
differential equation:
Fig. 4 Simulation result
1 § we we we · 21 Q § wT wT wT ·
2 u , v, w ¨ , , ¸ D¨ , , ¸
1 2Q ¨© wx wy wz ¸¹ 1 2Q ¨© wx wy wz ¸¹ The penetration depth and the displacement were compared
wu , wv , ww between the simulation and the fundamental experiment, as
e Hx Hy Hz, Hx H Hz (2)
wx y wy wz shown in Fig. 5 and 6.
The simulation results of the bead width and the penetration
where QandD are Poisson's ratio and the coefficient of
depth was agreed with the fundamental experimental result.
linear thermal expansion.
In the fundamental experiment, the displacement was
The authors tried to denote the latent heat in the
measured with a laser displacement meter. The behavior of
numerical simulation, i.e, the specific heat is changed
the displacement at the point a on the base metal, as shown
between 1807 K and 1809 K to denote transition
in Fig. 1, is shown in Fig. 6. A good agreement between the
temperature between solid and liquid phase. The parameters
simulation and the experiment was obtained.
were listed on Table 1.
Moreover, the other experiment with other welding
conditions was carried out to confirm the validity of the
4. Boundary conditions in numerical model
numerical simulation. The results of simulation and the
The Newton cooling law was applied on the surface
experiment are shown in Fig. 7 and 8, respectively.
except for that in contact with the carrier.
wU 6. Conclusions
Uc q h(u 293) (3)
wt The numerical model was designed to investigate the
Since the heat capacity of the carrier is very big, the dynamic behavior of the displacement due to the welding
temperature for the surface in contact with the carrier is distortion. In spite of the difference between the numerical
fixed to 293 K corresponding to the atmosphere simulations and the experiments about the temperature
temperature. The heat input to the base metal is divided to behavior on the back side of the base metal, the behavior of
two parts, to be agreement with the penetration of the weld the displacement in simulation due to the welding distortion
pool, as shown in Fig. 3. One is the heat input for the base almost agreed with experiment.
metal. Other one is heat input for the weld pool. If the penetration shape in the numerical simulation is
The heat input is denoted by the following equation: the same as the experiment, the welding distortion can be
predicted.
3Q 3x 2 3z 2
q1 ( x, y, z ) 2
exp( ) (4)
) exp(
ȧr r2 r2
where r is the radius of the heat input to be 6.6 mm.
5. Simulation result
Numerical simulation was carried out by FEM, for
which the software is Mentat 2007. According to the
movement of the heat, the weld pool corresponding to a
gray color is moved and the deposited metal was made, as Fig. 5 Comparison of penetration shape
shown in Fig. 4(a). The displacement is denoted as shown in fundamental experiment
191
Transactions of JWRI, Vol.39 (2010), No. 2
192
Transactions of JWRI, Vol.39 (2010), No. 2
193
Measurement of dynamical variation in two-dimensional temperature distribution of TIG pulsed-arcs
㸦A㸧 㸦B㸧
Fig.3 Experimental result (a) and simulation result (b) of 㸦C㸧 㸦D㸧
two-dimensional temperature distributions
Fig.6 Time variation of temperature distributions of TIG
T(K) T(K) pulsed arc at the times (A) ~ (D) marked in Fig. 5
25000 25000
22500
20000
22500
20000
4. Conclusions
17500 17500 The dynamic variation in two-dimensional temperature
15000
12500
15000
12500
distribution of a TIG pulsed-arc was measured through the
10000 10000 Fowler-Milne method with a high speed video camera as a
7500
5000
7500
5000
first step of the study. Consequently, it was found that the
arc column was expanded in the radial direction and the
1mm 1mm maximum arc temperature was 20,000K during the peak
current of 200A. On the other hand, the width of the arc
column decreased especially in the downstream region of
the arc and the maximum arc temperature fell to 17,500K
during the base current of 50A.
References
[1] G N Haddad and A J D Farmer: J. Phys. D:Appl. Phys. 17
(1984), 118
䠄a䠅㻌 㻌 㻌 㻌 㻌 㻌 (b)
㻌
Fig.4 Comparison of temperature distributions between [2] Hiraoka,et : Quarterly journal of the Japan Welding Society,
14 (1996), pp.641-648
cases of conical tip 45°(a) and 60°(b)
[3] M.Ushio, M.Tanaka and J.J.Lowke: IEEE Tran. P. S., 32
(2004), pp. 108~117
[4] Hiraoka et: Effect of electrode geometry on maximum arc
pressure in gas tungsten arc welding. Data of welding society
in Japan.
194
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Plasma MIG welding) (Bead formation) (Pure argon atmosphere)
1. Introduction
In the plasma MIG welding process, because shielding
gas is ionized in advance by a plasma electrode before
supply, the shape of arc is easily controlled through
electromagnetic force. Therefore the controllability of the
arc is remarkably enhanced compared with that in a
conventional MIG welding process. Because of this
advantage, it can be employed for MIG welding in pure
inert gas atmosphere. Although experimental results on
plasma MIG welding in pure argon atmosphere were
reported, further improvement of the welding process is
required because it is difficult to form a bead stably due to
lack of the wettability [1]. We have developed a new torch
for plasma MIG welding which has a shorter distance Fig. 1 Schematic illustration of Plasma MIG welding
between a contact tip and the top of a nozzle than that of the
conventional torch. In this study, experimental results on Contact tip
welding of steel plates with V groove in pure argon
atmosphere employing the new torch are discussed Plasma Electrode
compared with those for the conventional torch. Plasma nozzle
195
Improvement of bead formation in plasma MIG welding process in pure argon atmosphere
Acknowledgement
Part of this study was carried under the research activities
Temperture[K]
of the Consortium, which are financially supported by
NEDO.
References
[1] T. Kozuru et al: Preprints of the national meeting of J.W.S.,
83 (2008) pp.330-331.
Weld pool
196
Transactions of JWRI, Vol.39 (2010), No. 2
197
Mechanism of undercut in high speed welding based on moveless TIG welding
Under cut depth Hc/mm
$6 $6 $6 $6 Fig
4 Schematic diagram of flow in welding pool
Heat input distributiong
The filling metal is another factor to effect undercut.
Fig3 Undercut depth of different welding
When other conditions have no change, the depth of
current under the same input energy
undercut will decreases along with the quantity of filling
metal increase, and will disappear finally, as shown in
time, during which the surface tension gradient will
Table 1. This is mainly because the addition of filling metal
increase, which enhances the trend that liquid metal flows
makes trend of flow from periphery to welding pool center
from periphery to welding pool center, so the depth of
lower, which will counteract the effect of liquid metal flow
undercut increases, as shown in Fig. 3.
to inner absolutely, when its quantity reaches a definite
As described in Fig. 4, the molten angle is expressed
value. Therefore, undercut can be restrained.
byȘ. A liquid element near the periphery whose volume,
Table 1 Quantity of filling metal
of which the volume is V㸪is chosen in this article. / R is the Mass of filling Depth of undercut
outflow of liquid metal, and / L is the inflow of liquid metal. metal㸦mg㸧 㸦mm㸧
When other conditions are the same, the lower the Șis ,the 0 0.1
higher the Li will be. On one hand, it is because the 30 0.066
distance between / R and / L decreases with the decrease of 70 0.018
Ș, which makes the two reverse liquid flows that, effected 100 0
by the viscous force of liquid and restricted by each other, 150 0
and the liquid canÿt reach the region near the periphery
easily. On the other hand, this is also because the cooling 3. ConclusionV
V
velocity of the welding pool is higher in the condition of To sum up, the existance of surface active elements,
smaller Ș , which makes the columnar crystals grow up temperature gradient, molten angle and filling metal, all of
quickly from periphery to form a mushy region near the those have important effects on undercut. When the content
periphery area , and the flow of / L is blocked. Both of them of a surface active element reaches definite values (e.g the
will cause the deficiency of liquid near the periphery area, content of S is about 0.02-0.04% in steel), undercut will
so the trend of undercut is enhanced. occur. Temperature gradient makes the degree of undercut
It can be seen from the fore analysis, that increasing increase along with its augment. Increasing molten angle
molten angle can decrease the drag of liquid flow near the can restrains undercut efficiently. Adding the quantity of
periphery area efficiently, and can slow down the freezing filling metal can also reduces the degree of undercut.
rate, which is favorable for the metal of welding bottom to
supply the area of periphery. So this has obviously effect to References
[1]Seigo Nishikawa. High Speed Arc Welding, Industrial
restrain undercut.
Robot,1999,26(5)㸸371-376
[2]すṊྐ.: ⁐᥋Ꮫᚿࠋ1982 ᖺ 30㸦8㸧㸸pp68-74
[3]G.M.Oreper,J.Szekely: Journal of Fluid
Mechanism,1984,147,pp53-79
[4]R.T.C. Choo, J. Szekely. Welding Research Supplement
1992,(3):77-93
[5]R.T. C.Choo, J. Szekely, R.C. :Welding Research Supplement
September 1990:346-361
198
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Metal transfer model) (Arc discharge model) (Unified model) (Fe-vapor) (Electro-magnetic
force) (Shielding gas)
metal vapor evaporates from a weld pool and metal drops, J gas q gas
Tbboundary
and run into arc plasma. The metal vapor makes the arc Liquid Liquid
Liquid o u n d a ry
199
Numerical analysis the influence of arc shape on metal transfer process
0 0
0 5 0 5 0 5 0 5 0 5 0 5
mm mm
(a) Temperature (a) Temperature
mm
5.0108A/m2 mm 5.0108A/m2
5
5
0ms 4ms 8ms 0ms 4ms 9.3ms
8 8 8 8 8 8
2.3x10 3.6x10 7.1x10 2.3x10 4.4x10 3.4x10
2 2 2 2 2 2
A/m A/m A/m A/m A/m A/m
2 2
2
0 2 0 2 0 2 0 2 0 2 0
mm mm
(b) Current density (b) Current density
Fig. 3 The time-change of the arc temperature and current Fig. 4 The time-change of the arc temperature and current
density distribution in 100%Ar density distribution in 90%Ar+10%Fe
200
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (In-Process monitoring) (Adaptive control) (Titanium) (Micro laser welding) (Butt seam welds)
1. Introduction
Micro butt-welding with a pulsed laser has been often
used for sealing batteries in automobile and electronics
industries [1,2]. A gap in such laser welding is one of the
most important manufacturing problems, because the gaps
may cause fatal welding defects such as shallow
penetrations with underfilling or non-bonded joints.
In-process monitoring and adaptive control has been
proposed as one of the useful procedures to stably produce
sound penetration welds without welding defects.
Recently, some articles have been devoted to the researches Fig. 1 Schematic experimental set-up of in-process
of in-process monitoring and advanced adaptive control monitoring and adaptive gap control in micro butt laser
welding
technology in laser welding [3, 4]. The authors [3] have
shown that the heat radiation signal levels increase in gas. The beam was focused into a I 150 Pm spot diameter
proportion to the molten pool diameters and accordingly the as shown in Fig. 1. The laser oscillator has the noteworthy
shear strengths of lap welds in thin A3003 aluminum alloy feature of 50 W maximum average output power, with peak
sheets, and demonstrated that adaptive control on the basis power changeable within 5 kW at intervals of 100 Ps
of the heat radiation signal could stably produce sound according to the external voltage signal. The reflected
welds or on-site repaired welds. light and heat radiation from the laser irradiated area were
In this research, micro butt welding of pure titanium monitored coaxially with the Nd:YAG laser beam. The
was used with a pulsed fundamental Nd:YAG laser beam, in-process monitoring signals were measured by pin photo
and the reflected light and the heat radiation from the laser- diode sensors. High-speed pictures of the molten area
irradiated area were measured as in-process monitoring during laser irradiation were taken at the frame rate of
signals. These in-process monitoring signals were 20,000 frames/s from the angle of 45 degrees under the
evaluated by the correlation with the spot diameter of a illumination light of a 22-mW He-Ne laser. As for
molten pool, the penetration depth or the gap. Moreover, adaptive control, the laser peak power was controlled at
the laser peak power was controlled on the basis of both the 150 Ps intervals during irradiation according to the
reflected light and the heat radiation detecting the gap and reflected light or heat radiation to produce designed-size
the spot diameter of a molten pool in order to suppress spot diameters with suppression of underfilling. The 150
underfilling and small spot diameters for, not only spot Ps-rapid adaptive control loop consists of 100 Ps
welding, but also seam welding with the pulsed laser. monitoring periods and 50 Ps for the estimation of laser
welding conditions on the basis of in-process monitoring
2. Materials and Experimental Procedures signals and the change in the laser peak power.
The material used was commercially available pure
titanium of more than 98 % in purity. The samples were 3 3. Experimental Results and Discussion
mm thick and 1 mm wide as shown in Fig. 1. In-Process Monitoring of Reflected Laser Beam and Heat
Micro butt welding was carried out with the pulsed Radiation
fundamental Nd:YAG laser in 40-L/min Argon shielding An example of typical monitoring results of reflected
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* Graduate School, Osaka University, Osaka, Japan Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
** JWRI, Osaka University, Osaka, Japan Japan
201
In-process monitoring and adaptive control for micro welding of titanium
202
Transactions of JWRI, Vol.39 (2010), No. 2
4. Conclusions References
The gap is one of the most important issues to be solved [1] T. Sakurai and Y. Nakagawa, Proc. 62nd Laser Materials
Processing Conf. Osaka, (2003), pp 102-110.
in laser welding of a micro butt joint, because the gap leads [2]H. Fukuda, K. Ito, H. Tmaki, Y. Yamauchi and T. Amazutumi,
to welding defects such as underfilling or a non-bonded SANYO Technical Rev., 28 (1995), pp 112-118.
joint. The experimental studies indicated that the adaptive [3] Y. Kawahito and S. Katayama, J. Laser Applications 17
control of the laser peak power on the basis of in-process (2005), pp.30-37.
[4] C. Begger and F. O. Olsen, J. Laser Applications 15 (2003),
monitoring could reduce the harmful effects due to gaps in pp.19-24.
micro butt laser welding with a pulsed laser beam.
203
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Arc welding process) (Two-Electrodes TIG Arc) (3D numerical model) (Heat input density)
(Arc pressure) (Total heat input)
z [mm] z [mm]
15 15
10000 K
10000 K
Fig. 2 Influence of electrodes distance on temperature and velocity field of arc plasma
204
Numerical analysis on heat source characteristics of two-electrodes TIG arc
205
Transactions of JWRI, Vol.39 (2010), No. 2
4. Conclusion References
In this work the heat source characteristics of two- [1] T. Ueyama, T. Uezono, T. Era, M. Tanaka, K. Nakata, Science
electrode TIG are numerically investigated with three- and Technology of Welding & Joining, 14, 4, (2009), 305-314
dimensional TIG arc model. Results obtained are [2] X. Leng, G. Zhang, L. Wu, Journal of Physics D: Applied
summarized as follows: Physics, 39, (2006)
(1) In the two-electrode TIG arc, both distributions of heat
input density and arc pressure significantly change
depending on electrode distance.
(2) The total heat input of the two electrode TIG arc does
not change so much even if the electrodes distance
changes.
(3) The two-electrode TIG arc with appropriate electrodes
distance serves as a “soft” heat source with high heat
input.
206
Transactions of JWRI, Vol.39 (2010), No. 2
†
Fundamental research on micro discharge process
Research on discharge of sub-millimeter size process
KEY WORDS: (Precise materials processing) (Glow discharge) (Arc discharge) (Glow-arc transition)
1000 600
800 500
Power 600 400
Oscilloscope 300
CCD camera Microscope supply 400 Voltage
200
200 Current 100
0 0
nozzle
VTR PC 0 2 4 6 8 10 12 14
Base metal Discharge time(ms)
207
Fundamental research on micro discharge process
V o lta g e ( V )
Discharge time :8ms
Base metal : Stainless steel
900
700
500
0 100 200 300 400 500
Current (mA)
Fig. 3 Appearance of Ar DC discharge-plasma Figure 4 indicates the melting spot at a base metal
surface formed with use of the nozzle of diameter of 0.3
constricted by the nozzle of diameter of 0.3 mm.
mm. The size of melting spot was formed with about
ίGap length 5mm, Discharge current 390mA, 0.3mm diameter. The center of melting spot is seen with
Discharge voltage 800Vὸ small dimple shape as seen in Fig. 4. This phenomenon
mainly occurred due to the high-current density at the
center of the melting spot, high arc-pressure, and local
evaporation of melting metal.
Figure 5 shows a relationship between current and
voltage of the discharge at atmospheric pressure. In the
figure average voltage and range of voltage fluctuation are
indicated. The average voltage gradually decreases with
increase of current beyond around 300mA. This is
considered to be arc-glow transition region.
G
4. Conclusion
The possibility of precise materials processing by
electrical discharge energy at glow region or glow-arc
transition region was experimentally investigated using a
new type nozzle which was designed to establish very
small diameter plasmas and an experimental power source
which can supply a wide range of voltage. The results
obtained are summarized as follows:
(1) When discharge current increases, discharge mode
changes from normal glow to abnormal glow, that is,
glow-arc transition region.
0.5mm (2) Sub-millimeter sized melting spots were formed at the
base metal surface using the constricted nozzle of 0.3
mm in diameter.
208
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Magnetic control) (Permanent magnet) (Arc plasma) (Weld penetration) (3D analysis)
1. Introduction
Magnetic field
Arc plasma is affected by external magnetic forces.
From this view point, conventional studies have been Arc plasma
conducted on controlling the arc plasma by magnetic force
N S
for effective welding [1-4]. We focused on the cusp type
of magnetic field, which changes the cross section of the arc
plasma from a circular to an elliptical shape. In the
previous study, the cusp magnetic field suggested by Maruo
et al. was given by solenoid coils [4]. On the other hand,
we made the magnetic field by permanent magnets. Its S N
characteristics are small device size and no other power Electromagnetic
source. We have applied this magnetic control to TIG arc pinch force Magnetic poles
welding and evaluated its performance experimentally and
theoretically. It was confirmed experimentally that the (a) Conventional TIG arc (b) Magnetized TIG arc
permanent magnets installed in the TIG torch can make an
arc plasma with an elliptical shaped cross section. As a Fig. 1 Principle of elliptical arc shape by cusp type
result, bead on plate welding with the magnetized TIG arc magnetic field
provided sound weld beads and the limit of welding speed
without a weld imperfection was expanded [5, 6]. Our
previous results show that arc
shape control by permanent S N
magnets is useful. Tungsten
electrode
2. The Aim of This Study
13 mm
Figure 1 shows the principle
of the elliptical arc shape
generated by the cusp type Arc plasma
magnetic field. A conventional
13 mm
TIG arc plasma is illustrated in N S
209
Influence of magnet configurations on magnetic controlled TIG arc welding
Hm = 5 mm Hm = 3 mm
Weld penetration
Crater Crater Bead appearance
D: depth, W: width [mm]
D = 2.4
magnets
Without
W = 9.9
D/W = 0.24
D = 2.1
Type A
W = 4.2
D/W = 0.49
D = 3.1
Type B
W = 7.1
D/W = 0.43
Stationary arc
(1 sec. after arc ignition) Bead on plate welding (10 cm/min)
210
Transactions of JWRI, Vol.39 (2010), No. 2
YU Chun *,
LU Hao *
1. Introduction
Electromigration (EM) is a directional mass transporting
phenomenon, which is induced by a high current density. It
is a critical reliability issue in electronic packaging, since
the dimension of the interconnects and solder bumps
continuously decrease, and meantime, the current stressing
increases. Up to date, there are a lot of reports on the
investigation of EM mechanism in the flip chip solder joints
from an experimental view [1,2]. A few works have also Fig. 1 Cross section of solder joint
been performed to study the current density distribution in
the solder bump by employing finite element modeling 3. Results and Discussions
(FEM). However, the concentration distribution of the Figure 2 shows the concentration fields of the Cu and Sn
elements in the solder bump in the EM process is unclear. elements, just under the driving force from chemical
This paper aims at investigating the concentration field in potential. It is seen that the diffusion is uniform.
the solder joints under current stressing. And the model can
be used to predict the EM lifetime of the solder joints.
2. Model
There is a well established function, which describes the
relationship among the atomic flux and different physical (a) (b)
field, it is,
* * * * * Fig. 2 Diffusion under chemical potential. (a) Cu; (b)Sn
J J CH J EM J TH J ST
Figure 3 reflects the cross-sectional views of the current
CD * * 㸦1㸧
( DC ) eZ j density distribution in the solder joints The current density
kT distribution is non-uniform in the XY plane; most current
CQ * D Cȍȍ concentrates on the entry and exit points inside the bump.
T ı H
kT 2
kT The current crowding at the cathode side induces a
According to the second Fick’s law, the concentration maximum current density, which occurs at the entry point
equation for a multi-field EM process is established, near the top Cu line.
wC CD * *
[( DC ) eZ j
wt kT 㸦2㸧
CQ * D Cȍȍ
T ı H
kT 2 kT
*
For simplification, J*TH , and J ST are neglected in this
work. Figure 1 shows the geometry model and mesh for a
Cu/Sn-0.7Cu/Cu joint. 2.85A current was applied.
211
Visualization of EM process by FEM
Concentration (wt%)
0.55
concentration field, as shown in Fig. 4.
0.50 away from the crowding area
0.45
0.40
0.35
0.30
0 20 40 60 80 100
Time (hr)
Fig.4 Diffusion of Cu under current stressing Fig. 6 Concentration change of Sn in cathode IMC
212
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Lotus-type porous metal) (Laser welding) (Anisotropy) (Thermal diffusivity)
ᴾ ᴾ ᴾ ᴾ ᴾ ᴾ ᴾ ᴾ ᴾ (Laser absorption coefficient)
1. Introduction
Lotus- and Gasar- type porous metals had been
developed by Boiko et al. [1, 2] and Nakajima et al. [3-5]
and these were expected as innovative engineering
materials because the directional pore yields various unique
properties [6]. Laser weldability of the lotus-type porous
copper [7], iron [8] and magnesium [9] was investigated,
and the effect of pore direction on laser fusion zone shape
of the magnesium [10-12] and the copper [11,12] by using
the results of 3D FEM analysis of temperature distribution
during welding were demonstrated. These have pointed out
that the relation between direction of the pores and the laser
irradiated direction appreciably influence weld formation.
In the present paper, we performed 3D FEM analysis of Fig. 1 Schematic views of the lotus-type porous metals
temperature distribution during laser welding for the lotus- showing combinations of pore direction, applied
type porous iron and compared the fusion zone shape with heat source direction, and welding direction.
those cross sections obtained by experiments [8]. We also
estimated the anisotropy of thermal diffusivity inherent in
the lotus-type porous metals, and the anisotropy of the laser
absorption coefficient caused by the phenomenon of
multiple reflections of laser on the wall of pores.
213
Numerical simulation of fusion zone shape of lotus-type porous metals produced by laser welding
3. Numerical simulation and discussions determined as shown in Table 1. There is very little
3D FEM calculations of temperature distribution of the difference between E and E ' .
Lotus iron were performed using ABAQUS with user-
defined subroutines. The Lotus iron is modeled as an
equivalent orthotropic material. Equivalent density, 4. Conclusions
equivalent specific heat, and thermal conductivity along and The conclusions of this study are summarized as follows.
normal to the directional pores with temperature (1) Fusion zone shape of the weld bead has little difference
dependencies are described by the pore volume content for three combinations. Calculated shapes and the
ratio H and the property of non-porous iron (AISI1006) experimental ones were similar.
[20]. (2) The difference between the equivalent thermal
Fusion shape of the weld metal is estimated by the cross diffusivity along and normal to the directional pores is
section of maximum temperature exceeding the melting small for the Lotus iron.
(3) The little difference of laser absorption coefficient E
and E ' for the Lotus iron.
Acknowledgment
This research was supported as part of the entrusted
project “development of lightweight high stiffness structural
materials and evaluation technology” for the “advanced
Fig. 3 Maximum temperature distributions of Lotus iron in machining system development project” in the fiscal year
the cross section at half of the welding (x) direction; 2005 consigned by NEDO. The authors gratefully
absorbable laser power of 0.522 kW, laser spot appreciate this support.
diameter of 1.0 mm, and welding speed of 1 m·min-1.
point at half of the x direction. Figure 3 shows maximum References
temperature distributions of the Lotus iron. The absorbable [1] L.V. Boiko et al.: Sov. Powder Metal. Met. Ceram., 30
laser power is assumed to 0.522 kW as the most similar (1991), pp. 78-81.
results between the experimental fusion depth and the [2] L.V. Boiko: Mater. Sci., 36 (2000), pp. 506-512.
calculated ones for the Combination 2 A A . Fusion [3] S.K. Hyun et al.: Proc. Int. Conf. Solid-Solid Phase
Transformations ’99 (JIMIC-3), (1999), pp. 341-344.
zone shape has little difference, and the calculated shapes [4] H. Nakajima et al.: Coloids Surfaces A, 179 (2001),
and the experimental ones were similar. pp. 209-214.
The reason is considered as follows; Equivalent thermal [5] S.K. Hyun et al.: Mater. Sci. Eng. A, 299 (2001),pp.241-248.
diffusivity along and normal to the directional pores are [6] H. Nakajima: Prog. Mater. Sci., 52 (2007), pp. 1091-1173.
described as follows: [7] T. Murakami et al.: Mater. Sci.Eng.A,357(2003),pp.134-140.
[8] H. Yanagino et al.: Mater. Trans., 47 (2006), pp. 2254-2258.
D eq// D n , D eqA 1 H D n
1
(1).
[9] T. Murakami et al.: Mater. Sci.Eng.A,456(2007),pp.278-285.
Table 2 shows the equivalent thermal diffusivity for the [10] T. Tsumura et al.: Mater. Sci. Forum, 502 (2005), pp. 499-
used metals. The difference between these values is small 504.
for the Lotus iron. [11] T. Tsumura et al.: Mater. Trans., 47 (2007), pp. 2248-2253.
Anisotropy of the laser absorption coefficient caused by [12] T. Tsumura et al.: Soild State Phenomena, 127 (2007),
the phenomenon of multiple reflections of laser on the wall pp. 307-312.
[13] S.K. Hyun et al.: Mater. Sci. Eng. A, 340 (2003), pp. 258-
of open pores is also considered; Number of multiple
264.
reflections n is described by: [14] T. Ogushi et al.: J. Appl. Phys., 95 (2004), pp. 5843-5847.
n int t tan T / d (2). [15] T. Ikeda et al.: J. Jpn. Foundry Eng. Soc. 74 (2002),
Here, t is sample thickness, T is laser irradiated angle, pp. 812-816 (in Japanese).
[16] T. Ikeda et al.: J. Jpn. Inst. Light Met. 54 (2004), pp. 388-
d is average pore diameter. Equivalent heat input along
393 (in Japanese).
and normal to the directional pores are described as [17] Z.K. Xie et al.: Jpn. J. Appl. Phys. 43 (2004), pp. 7315-
follows: 7319.
Q Ain = E Q, Q in// = [(1-H )E + (1-(1-E )n )H ]Q=E ' Q (3). [18] M. Tane et al.: Acta Mater. 52 (2004), pp. 5195-5201.
[19] E.A. Brandes: Smithells Metal Reference Book, 6th ed.,
Therefore, laser absorption coefficient E and E ' are Butterworths, London, (1983), p. 17-7.
[20] M.R. Frewin et al.: Welding Research
Table 2 Estimated thermal properties of material used at room temperature. Supplement, (1999), pp. 15s-22s.
[21] Japan Society of Thermophysical
Properties Ed.: Thermophysical
Properties Handbook, Yokendo LTD,
Tokyo, (1990), p. 23 (in Japanese).
214
Transactions of JWRI, Vol.39 (2010), No. 2
Numerical simulate on the coupled arc and pool for GTAW using a
†
unified mathematical model
LEI Yongping*, SHI Yaowu*, LIN Jian *, LU Zhenyang **, XIAO RONGSHI*** and
MURAKAWA Hidekazu****
KEY WORDS: (Numerical simulation ) (Unified mathematical model) (Coupled arc and pool ) (Boundary-
fitted coordinate system ) (GTA) (Welding)
215
Numerical simulation of the coupled arc and pool for GTAW using a unified mathematical model
temp temp
2 20000 2 20000
19000 19000
18000 18000
4 17000 4 17000
16000 16000
15000 15000
6 14000 6 14000
13000 13000
12000 12000
z /mm
z /mm
8 11000
8 11000
10 10
12 12
L = 5 mm
14 t = 0.5 s 14 L = 5 mm
t = 1.0 s
16 16
-10 -8 -6 -4 -2 0 2 4 6 -10 -8 -6 -4 -2 0 2 4 6
r /mm r /mm
(a) heating time is 0.5 second (b) heating time is 1.0 second
temp temp
2 20000 2 20000
19000 19000
18000 18000
4 17000 4 17000
16000 16000
15000 15000
6 14000 6 14000
13000 13000
12000
z /mm
12000
z /mm
8 11000 8 11000
10 10
12 12
14 L = 5 mm 14 L = 5 mm
t = 1.5 s t = 2.0 s
16 16
-10 -8 -6 -4 -2 0 2 4 6 -10 -8 -6 -4 -2 0 2 4 6
r /mm r /mm
(c) heating time is 1.5 second (d) heating time is 2.0 second
Fig.2 Weld arc temperature distribution and weld pool shape and size under different heating time
seen that the volume of the weld pool is increasing with the
welding time extending, but the change of arc temperature 4. Conclusions
distribution is slight. (1) The developed model is suitable to describe the
Figure 3 shows the temperature fields in two kind of behavior of the integrated welding arc and pool system.
shielding atmospheres. It can be see that the highest (2) The two-way interaction of the arc and pool
temperature decrease and the high temperature area is subsystems can be calculated by the two sub-model
contracted into the arc center. with “clear” interfaces.
216
Transactions of JWRI, Vol.39 (2010), No. 2
(3) The influence of weld pool deformation on the arc supplying the thermo-physical properties of the argon
temperature fields is very small when the welding plasma.
current is 200A.
References
Acknowledgment [1] Lowke J.J., Proceedings of International Symposium on
This work was supported by the Beijing Nature Science Theoretical Prediction In Joining and welding, Editor
Foundation under grant No.3102002 which is gratefully M.Ushio, JWRI, Osaka University, 1996, 1-9
acknowledged. The author also thanks Prof Murphy A.B, [2] Choo R.T.C., Szekely J., Westhoff R.C., Metall. Trans.
CSIRO Division of applied physics, Australia, for his B ,1992, 23B: 357-369
[3] Lei Y.P., Gu X.D., Shi Y.W. and Murakawa H., Acta
Metallurgica Sinica (Chinese), 2001, No.5, 537-542.
217
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (475°C embrittlement) (Austenitic stainless steel) (Spinodal decomposition) (Phase field model)
(Prediction of hardening) (Computer simulation) (Solidification mode)
1. Introduction
It has been widely recognised that ferritic stainless
steels with more than 20mass%Cr are embrittled by
exposure to temperature of 673-823 K known as “475°C
embrittlement”. This embrittlement is generally
attributed to the separation of an originally single phase
to one containing an iron-rich bcc phase (D) and a
chromium-rich bcc phase (D’) caused by spinodal
decomposition. On the other hand, the chemical
compositions of the austenitic stainless steel weld metals
have been designed to contain the 3-10vol% ferritic
phase in order to reduce the hot cracking susceptibility.
The objective of the present study is to validate the
possibility of spinodal decomposition in the austenitic
stainless steel weld metal. The effect of alloying
elements in a ferritic phase, which is influenced by the
solidification mode, on the behaviour of spinodal Fig. 1 Comparison of microstructure in weld metal
between SUS316L-1 and SUS316L-2
decomposition was investigated by the computer
simulation based on the phase filed model for ternary Table 2 Analysed compositions of G-ferrite (at%)
alloy system.
218
Prediction of 475°C embrittlement in stainless steel welds using phase field model
4. Computer Simulation of Spinodal Decomposition (D) and Cr-rich (D’) phases with the progress of ageing.
Phase Field Model of Spinodal Decomposition for Ternary In the initial stage of spinodal decomposition, Mo is
Alloys distributed along with Fe in the D phase. Then, the
The evaluation equations governing spinodal bifurcation of Mo peaks along peak tops of Cr
decomposition in the A-B-C ternary system are concentration occurs with the progress of phase
expressed as:
GG sys GG sys
decomposition. Then, Cr-rich D’ phase grows and
wc 2
wt
= L 22
Gc 2
+L 23
Gc 3
coarsens as well as slightly modulating along <100>
(1) direction with the lapse of ageing time. Mo is finally
wc 3 GG sys GG sys
wt
= L 32
Gc 2
+L 33
Gc 3
redistributed to the interface between D and D’ phases,
and surrounds Cr-rich D’ phase.
where c i is the local composition of the component i at The effects of Cr and Ni contents in Fe-Cr-Ni ternary
spatial position r (r represents a vector) and time t in the alloy on the beginning time for spinodal decomposition
microstructure. The subscript numbers i = 1, 2 and 3 were investigated. Figure 3 shows the predicted contour
refer to elements A, B and C, respectively. G sys is the of the beginning time for spinodal decomposition as the
total free energy of the microstructure, which is function of Cr and Ni compositions. The beginning time
expressed by the sum of the Gibbs energy density, the for spinodal decomposition shortens with increasing Cr
density of the composition gradient energy and the
density of the elastic strain energy. L ij is the Onsager
coefficient.
In the present study, it is assumed that the change in
hardness due to spinodal decomposition and/or 475°C
embrittlement is followed by the precipitation hardening
theory (cut-through model) proposed by Mott-Nabarro.
219
Transactions of JWRI, Vol.39 (2010), No. 2
5. Conclusions
(1) The original single phase was phase-decomposed into
Fe-rich and Cr-rich phases with the progress of ageing.
The beginning time for spinodal decomposition
shortened with increasing Cr composition, while it
much lengthened contrarily with increasing Ni
composition.
(2) The predicted hardness of a ferritic phase with FA mode
began to increase as the holding time approx. 5 Gs,
while that with AF mode began to increase at the
holding time as long as approx. 150 Gs.
220
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Alloy 690) (Multipass weld metal) (Microcracking susceptibility) (Impurity elements)
ᴾ ᴾ ᴾ ᴾ (Grain boundary segregation) (Molecular orbital calculation) (Cracking mechanism)
1. Introduction
A Ni-base superalloy, alloy 690 is highly susceptible
to ductility-dip cracking, and the primary cause of
ductility-dip cracking in the reheated weld metal of alloy
690 is likely to be the reduction of hot ductility
attributable to grain boundary segregation of impurity
elements such as P and S [1]. The objective of the
present study is to clarify the cause of the increased
ductility-dip cracking susceptibility with an increase in
the P and S contents based on a grain boundary
segregation analysis and a molecular orbital analysis of Fig. 1 Relation between (P+1.2S) in weld metal and
ductility-dip temperature range, DTR
the binding strength of the grain boundary.
2. Materials and Experimental Procedures compositional parameter of (P+1.2S) and the DTR for
The base metal used is a commercial alloy 690. all weld metals. According to the multipass weld
Several kinds of commercial filler metals (FM1-FM3), cracking test results, the critical DTR where microcracks
lab-melting filler metals (FF1-FF5) and extra high-purity didn’t occur during multipass welding could be
filler metal (EHP) were employed for comparison. The estimated as approx. 200K. It follows that the amount of
chemical compositions of base metal and filler metals (P+1.2S) in the weld metal should be limited to 30ppm
used are shown in Table 1. The ductility-dip cracking in order to prevent ductility-dip cracking in the multipass
susceptibility in the reheated weld metal was evaluated weld metal.
by the spot-Varestraint test.
4. Grain Boundary Segregation of P and S
Table 1 Chemical compositions of steels used (mass%)
Numerical Model of Grain Boundary Segregation
The numerical model of microsegregation involves
segregation during the solidification stage and
segregation/ desegregation during the cooling/reheating
stage following solidification in welding under the initial
221
Metallurgical mechanism of ductility-dip cracking in multipass welds of alloy 690
Fig.4 Schematic illustration of cluster model used Fig.5 Calculated bond order between Ni atoms for
(molecular orbital analysis by DV-XD method) different grain boundary segregation scenarios
222
Transactions of JWRI, Vol.39 (2010), No. 2
7. Conclusions
6. Mechanism of Ductility-Dip Cracking in Multipass (1) A numerical analysis of the segregation behaviours of P
Weld and S revealed that these elements were cosegregated at the
A grain boundary segregation analysis revealed that grain boundary during multipass welding. P and S
P and S (especially S) were segregated to the grain concentrations at the grain boundary slightly increased with
boundary in the weld metal during multipass welding. A an increase in the number of thermal cycles applied.
molecular orbital analysis suggests a possibility that (2) A molecular orbital analysis suggested a possibility that
grain boundary segregation of P and S leads to grain grain boundary segregation of P and S led to grain
boundary embrittlement. It follows that the ductility-dip boundary embrittlement.
cracking in the reheated weld metal of alloy 690 would
References
be dominantly caused by the embrittlement of grain [1] K.Nishimoto, K.Saida, H.Okauchi and K.Ohta : Sci. and
boundaries resulting from the imbalance between Technol. of Welding and Joining, 11-4 (2006), p.462-470.
intergranular strength and intragranular strength at high
temperature attributable to grain boundary segregation
of impurity elements such as P and S.
223
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Lead-free solder) (Intermetallic compound) (Reaction layer) (Under bump metallurgy)
(Co-P plating) (Ni-Co-P plating)
1. Introduction the center of the substrate and activated flux (0.01 ml) was
In response to health and safety concerns, lead-free dropped on the solder. The test specimen was put into a
soldering has become a popular technology in electronics radiation furnace in a nitrogen atmosphere and heated
packaging. Compared with the lead-containing solders, Sn- according to the temperature rise profile shown in Fig. 2.
3.0mass%Ag-0.5mass%Cu (SAC, all mass% unless The reflow peak temperature was 513 K with the sample
specified otherwise) solder widely used in Japan has a above 490 K for 115 s. After soldering, the spreading area
relatively low impact reliability owing to the solder alloy of the solder on the UBM was measured by using the
hardness that induces a high stress concentration at the optical microscope (OM). Three tests were conducted to
interface. In general, there is a correlation between the obtain an average value for each specimen. Then,
impact reliability and the morphology and thickness of the specimens were cut and the cross-section of the specimens
reaction layer formed at the solder/under bump metallurgy was polished to observe the interface between the solder
(UBM) interface. The most common UBM is electroless and UBM. The reaction layer at the interface was observed
Ni-P plating over copper pad. Electroless Ni-P acts as a by scanning electron microscope (SEM).
diffusion barrier layer between the copper and the solder.
However, due to nickel diffusion, P-rich layers form at the 3. Result and discussion
interface between the solder and electroless Ni-P. Solder A spreading test was carried out to examine the
joint failure is related to the growth of these layers and to wettability of solders on UBMs and to clarify the interfacial
their brittleness and affects the mechanical reliability of reaction between the solder and UBMs after soldering. The
joints. Recently, a new composition of UBM is proposed as wettability of solder was experimentally assessed by
diffusion barrier1-4). For instance, Magagnin et al. reported measuring the spreading area of solders on the UBMs.
that electroless Co–P strongly limits interdiffusion and
intermetallic compounds formation as compared with the
electroless Ni–P with Sn-Ag-Cu alloy. Furthermore, in the
Co-P samples, P-rich layers did not form at the interface4).
It is important to investigate the relationship between
morphology of reaction layer and UBM. This study aims to
clarify the effect of Co-P and Ni-Co-P on the morphology
of reaction layer formed at the solder/UBM interface.
2. Experimental
SAC solder (0.3 g) was used in this study. Electroless
Co-P(Au) (3.1 Pm) and electroless Ni-Co-P(Au) (5.2 Pm)
finished Cu plates on FR-4 PCBs (25.0×25.0×1.6 mm)
were prepared as UBM. Electroless Ni-P(Au) (5.0 Pm)
substrate was also used as a reference substrate. These
substrates were plated with gold to avoid oxidation of the
cobalt and nickel surface.
The experimental procedure is shown in Fig. 1. The Fig. 1 Experimental procedure of this study.
substrate was immersed in 4% HCl solution for 120 s and
then rinsed with deionized water. Then, solder was put on
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* Graduate School, Osaka University, Suita, Japan Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
** JWRI, Osaka University, Ibaraki, Japan Japan
*** Okuno Chemical Industries Co., Ltd., Osaka, Japan
224
Interfacial reaction between Sn-3.0Ag-0.5Cu solder / Co-P plating and Ni-Co-P plating
References
[1] Chwan-Ying Lee et al., Thin Solid Films, 239, (1994), pp.93-
98.
[2] Yi-Hsiang Chao et al., Metallurgical and Materials
Transactions A, 39, (2008), pp.477-489.
[3] Takashi Yamamoto et al., Materials Transactions, 46, (2005),
Fig. 3 Appearance of typical spreading test specimen after pp.2406-2412.
[4] L. Magagnin et al., Electrochimica Acta, 50, (2005), pp.4621-
soldering. a) electroless Co-P, b) electroless Ni-Co- 4625.
P, c) electroless Ni-P
225
Transactions of JWRI, Vol.39 (2010), No. 2
TAKAHASHI Makoto *, WATATANI Yuuki **, IKEUCHI Kenji *, KIMURA Hisamichi ***
and INOUE Akihisa ***
KEY WORDS: (Anodic bonding) (Metallic glass) (Silicate glass) (Joinability) (Microstructure)
226
Formation of reaction layers at anodically-bonded metallc glass/silicate glass interfaces
5 mm 5 mm
anodically-bonded joint across the bond interface with the anodic bonding. At a higher temperature drift of alkali ions
focused ion-beam (FIB) system. The FIB system used was in silicate glass with voltage application is enhanced more
Hitachi FB-2000A, and TEM observation was performed strongly, and buildup of the alkali ion depletion layer and
by JEOL JEM-2010 operating at an accelerating voltage of consequently the bonding process become quicker. On the
200 kV. other hand, anodic bonding of metallic glass to 7622 glass
showed a peculiar progress (Fig. 1(b)). At T b = 563 K, the
3. Results and Discussions 7622 glass joints once achieved intimate contact of almost
In Fig. 1 are shown the appearances of the metallic all the joint interface with the voltage application for 300 s.
glass silicate glass anodically-bonded interfaces. The 801 But the joint interfaces were separated again with longer
joints shown in Fig. 1 were bonded at 563 K, and the 7622 voltage application. At a higher T b of 603 K, all the
joints were bonded at 563 K. In the joint interfaces, the metallic glass / 7622 glass joint interfaces remained
areas in which intimate contact of the metallic glass to the separated with the voltage application from 75 s to 300 s.
silicate glass showed a little dark tone, and the areas in Microstructures around the joint interfaces in a 801
which metallic glass was separated from the silicate glass glass joint anodically-bonded at 563 K with voltage
showed bright tone with Newton fringes. (In the 7622 glass application for 300 s and a 7622 joint bonded at 523 K with
joint of t b = 1200 s all the joint interface was separated. voltage application for 4800 s were observed by TEM. In
The appearance was oberved without 7622 silicate glass.) Fig. 2 is shown the microstructure of the joint interface in
As is shown in Fig. 1(a), in 801 glass joints the ratio of the the 801 joint. Reaction products formed two layers at the
intimate contacting area increased with the voltage joint interface. One is a crystalline layer ~70 nm thick
application time, and progress of the bonding became faster contacting the metallic glass, and another is an amorphous
at higher bonding temperatures. This is a usual tendency in layer ~20 nm thick contacting the silicate glass. Chemical
analysis by energy dispersion X-ray Spectroscopy (EDS)
(a) Metallic glass (b) showed existence of high-concentrations of oxygen in these
layers. Therefore, these reaction layers were thought to be
of oxide formed by reaction of elements from the metallic
0 glass with oxygen from the silicate glass in anodic bonding.
And Cu concentration in the amorphous oxide layer was
Reaction
801 silicate glass very low. The microstructure around the joint interface in
layers
227
Transactions of JWRI, Vol.39 (2010), No. 2
228
Transactions of JWRI, Vol.39 (2010), No. 2
1. Introduction
Recently, because of aluminum’s light weight, it has 3. Results
been considered an energy-saving structural material in At first Keller reagent is used to corrode the specimens,
advanced applications. In addition, aluminum is an easily second the specimens are washed by the 4% of the nitrate
saved resource because it can be recycled, and thus can be acid, at last, the specimens are observed by optical
expected to be an environmental friendly metallic material. microscope. FSW joint can be divided into the following
One such application would be its use in automobiles, several zones(see Fig. 1) : A: ḛ Half-leafḜ zone, B:zone
which would facilitate transportationύ numerous similar above WNZ, C:TMAZ, D:WNZ, E:boundary of AS,
examples could be cited in support of employing aluminum F:boundary of RS, G:HAZ, H:Base metal.
as a structural material[1]. In addition, because aluminum
has large thermal conductivity, large specific heat, quick
thermal conductivity and large linear expansion coefficient,
it easy to form an air hole during the welding[2]. So, welding
Aluminum welding is a difficult problem. Two kinds of
mature approach are Friction Stir Welding(FSW) and
Variable Polarity Plasma Arc Welding(VPPAW).
Therefore, it is important to compare the microstructure and Fig. 1 Macro Map of FSW joint
micro-hardness of the Friction stir welding joints and
Variable polarity plasma arc welding joints. The regions are observed under the microscope. And
Therefore, the friction stir welding and the variable the regions show different characteristics, and asymmetry
polarity plasma welding are compared, and the study is features of the FSW joints
based on differences of the microstructure and mechanical
properties of these two methods. A nano-indentation
experiment is used to test the hardness in some narrow and
small zones, such as the ḛHalf-leafḜzone . The nano-
indentation method is flexible and accurate and provides a
powerful tool to test the hardness of weld joints. It is the
basic work for the hybrid welding has laid a solid
ίaὸḛHalf-leafḜzone corroded ίbὸḛHalf-leafḜ zone corroded
foundation. by Kellar acid by aqua regia
Fig. 2 microstructure of ḛHalf-leafḜzone
2. Experimental procedures
Specimens of the FSW and VPPAW joints are Figure 2(a) shows the ḛHalf-leafḜzone, the zone is
obtained from 6mm 2219 aluminum alloy plate. By using under strong pressure. ḛ Half-leaf Ḝ zone can be shown
an optical microscope, the microstructure of welding joint clearly by use of aqua regia, and this area can be corroded
is observed. The procedures are sampling, polishing,
easier than other zone(see. Fig. 2(b)).
corrosion and observation. The cross sections of
metallographic specimens are polished, etched with
Keller’s reagent (1 ml hydrochloric acid,1.5 ml nitric acid,
PO K\GURÀXRULF DFLG DQG PO ZDWHU DQG washed by
4% of the nitrate acid. At last specimens are observed by
optical microscopy. WhDW’s more, aqua regia is used. By
using SEM, microstructure is very clearly seen. After these ίaὸḛHalf-leafḜzone ίbὸbound ofḛHalf-leafḜzone ίcὸnear theḛHalf-leafḜ
experiments, a Nano-indentation test is done. All the tests zone(TMAZ)
were conducted on a Nano Indenter G200. Fig. 3ᴾ result of ḛHalf-leafḜzone by Scanning electron
microscopy
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* Beijing University of Technology, China Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
Japan
229
The dissymmetry of friction stir welding joints and variable polarity plasma arc welding joints study
4. Conclusions
In this paper, two welding methods which are FSW
ίcὸHL ίdὸHAZ and VPPAW are used to weld 2219 aluminum alloy. Finally
both the FSW and VPPAW will appear non-symmetrical
welded joints, such non-symmetry is reflected in the
microstructure morphology and performance.
230
Transactions of JWRI, Vol.39 (2010), No. 2
(1)In the FSW joints, defects in onion rings, the VPPAW, VPPAW joint can be adjusted to symmetry and
different of forward side and backward side microstructure results of satisfactory welding. Not only symmetrical
and ḛHalf-leafḜzone have made FSW joints have a strong microstructure is obtained, but also good properties of
dissymmetry. The dissymmetry in welded joints is brought welded joints are obtained by using the right filling material.
about the dissymmetry in joint performance. The
dissymmetry in FSW joint can not be suppressed due to the References
nature of the mixing head and the process of FSW, it is [1] Tomotake Hirata ,Taizo Oguri, et al. Materials Science and
Engineering A456 (2007) 344–349
difficult to over come.
[2] Huang Wangfu. Aluminum and aluminum alloy welding
(2)The problems about dissymmetry are also present guide.Chang Sha: Hunan Science and Technology Press,2004
in the VPPAW joint. However by changing the form of
231
Transactions of JWRI, Vol.39 (2010), No. 2
,
THATER Raphael*, PERRET William**, SCHWENK Christopher* **, ALBER Uwe ***
,
and RETHMEIER Michael* **
KEY WORDS: (Numerical welding simulation) (Temperature field) (Distortion) (Industrial application) (Auto-
motive assembly)
232
Industrial application of welding temperature field and distortion visualization using FEA
proposed by Goldak [2] has been used to describe the MIG- fied. The welding simulation therefore offers significant
welding process. After calibration of this phenomenological benefits in terms of visualization in comparison to experi-
heat source, the simulated temperature distribution is in best mental methods. Nevertheless, an advantage of the experi-
agreement with experimental data. mental observation is the direct measurement of the real
The necessary temperature dependent material proper- component behavior. The result quality of a welding simu-
ties of an AlMgSi alloy were taken from [3]. Softening lation, however, is highly dependent on the accuracy of the
effects occurring in the heat affect zone of a T6 tempered calibrated temperature field and the chosen assumptions
aluminum alloy are taken into account using the micro- and simplifications.
structure model developed by Leblond [4]. The parameters The result accuracy of this study is illustrated in Fig. 3
of this model were calibrated against hardness measure- as a comparison of the numerical and experimental tran-
ments across a weld seam of the actual alloy. sient distortion behavior at point A. Qualitatively, the cha-
The calibrated temperature distribution is used as an racteristics of the welding distortions are well reproduced
input for the subsequent mechanical calculation, which by the simulation. Quantitatively, the calculated distortion
determines the distortions according to the calculated tem- is approximately 60 % higher than the measured one. Con-
peratures. The mechanical boundary conditions were ap- sidering the simplifications and assumptions of the simula-
plied to the experiments equivalent places and removed at tion model the result accuracy is reasonable.
equivalent times. Orthotropic elastic constraints were used
to reproduce the clamping conditions. It should be noted
that, in contrast to the production, in this study the assem-
bly was not fully unclamped. One side of the crossmember
was clamped over the entire measurement period to sup-
press rigid body motions and thus enable the direct compar-
ison of experimental and numerical results at defined refer-
ence points.
5. Conclusions
As shown in the previous section the increasing compu-
tational power enables us to run a 3D nonlinear transient
welding simulation for a large industrial geometry in a time
frame that allows a widely industrial application of welding
simulation. In this case, the calculation could be done over-
night on a desktop PC.
Fig. 2 Simulated distortion distribution (magnification fac- Finally the results of this study show that the calculation
tor: 25). of complex and large industrial parts with good result accu-
racy and a reasonable calculation time is possible. Welding
In contrast to that, the experimental visualization of simulation, therefore, has a great potential with its strengths
many of these physical data is very complex, restricted or in visualization and the specific examination of parameter
even impossible. Clamping devices, welding robots and the influences to reduce experimental effort within a target-
welding process itself limit an observation during the weld- oriented distortion optimization in industrial environments.
ing process to small sections. An experimental visualization However, additional research is needed with respect to the
of the entire assembly is usually possible only for un- time-to-solution (e.g. automatic heat source adjustment and
clamped conditions before and after the joining process. meshing) and the user-friendliness of the software in order
Furthermore, the numerical approach offers, in contrast to become even better suited to industrial requirements.
to experiments, the opportunity to examine the influences
of individual parameters by an uncoupled parameter varia- The conclusions of this study are summarized as follows.
tion, so that the fundamental mechanisms responsible for (1) A welded assembly from the automotive industry has
the deformations of welded parts can be clearly identi- been investigated with experimental and numerical
233
Transactions of JWRI, Vol.39 (2010), No. 2
methods. References
(2) The results of the simulations show that current re- [1] W. Perret, R. Thater, U. Alber, C. Schwenk, M. Rethmeier:
sources allow an industrial applied welding simulation Case study for welding simulation in the automotive industry,
with good result accuracy and reasonable calculation IIW-Doc. SC-Auto-37-10 (2010).
times. [2] J. Goldak, A. Chakravarti, und M. Bibby: A new finite element
model for welding heat sources, Metallurgical Transactions B
(3) Compared to experimental methods the flexibility of
15B (1984), pp.299-305.
welding simulation offers great advantages in terms of [3] SYSWELD™ 2009: Material Database, ESI Group.
visualization. Many physical data can be visualized [4] J. B. Leblond and J. Devaux: A new kinetic model for
without any problems and restrictions. anisothermal metallurgical transformations in steels including
effect of austenite grain size, Acta Metallurgica 32 (1984), pp.
137-146.
234
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Welding residual stress) (Electron beam welding) (Hot isostatic pressing) (Finite element
method) (Cooling channel) (Test blanket module)
235
Prospective design of weld joint between first and side walls in test blanket module for ITER
45
1490mm
40 FEM modeling Side wall
35
200mm
Side wall
30 Weld
First wall
422mm
x direction First wall
z [mm]
25
Case1 y Weld line
20 Case2 (180mm) 200mm
Case3 z 232mm 116mm
15
Case4
10
21.5mm
-400 -300 -200 -100 0 100 200 300 400 500 600 700
12mm
Residual Stress along Y-direction [MPa]
25mm
25mm
Fig. 2 Residual stress distributions parallel to weld line 10mm 16mm
beside cooling channels.
(Current Design) (Prospective Design)
3. Evaluation of Prospective Design Fig. 3 Schematic illustration of box structure for TBM and
Figure 3 shows a schematic illustration of the box FEM models with current and prospective design.
structure for Japanese TBM and FEM models for numerical
analysis. As shown in this figure, a partial model was Case-1 Case-2
x
employed for this evaluation because of the limitation of
Weld line
computer resources. This figure also shows the FEM model z y
with current and prospective designs obtained from the
above studies. In prospective design, the width of backing
16mm
was modified without changing the thickness of first wall. 10mm
In the order to evaluate the residual stress in the prospective
design, the thermal elastic-plastic analyses were conducted
in two cases, which were Case-1 (current design) and Case- Fig. 4 Plastic strain distributions parallel to weld line
2 (prospective design), respectively. at cross sectional plane of middle of weld line.
In the elastic-plastic analyses, the mechanical boundary
[MPa] Case-1 Case-2
conditions were set as follows.
1) Displacement along y-direction was fixed at the x
Weld line
cutting planes of the first and side walls
perpendicular to the welding direction z y
2) Displacement along z-direction was fixed at the
other cutting plane of the first wall
16mm
10mm
236
Transactions of JWRI, Vol.39 (2010), No. 2
around cooling channels was small and its distribution M.A. Sokolov, R.E. Stoller and S.J. Zinkle: Fusion
was more uniform than that in current design. Engineering and Design, 83, (2008), pp.1471-1476.
[2] T. Hirose, M. Enoeda, H. Ogiwara, H. Tanigawa and M.
References Akiba: Fusion Engineering and Design, 83, (2008), pp.1176–
[1] H. Tanigawa, T. Hirose, K. Shiba, R. Kasada, E. Wakai, H. 1180.
Serizawa, Y. Kawahito, S. Jitsukawa, A. Kimura, Y. Kohno, A. [3] H. Serizawa, S. Nakamura, M. Tanaka, Y. Kawahito, H.
Kohyama, S. Katayama, H. Mori, K. Nishimoto, R.L. Klueh, Tanigawa and S. Katayama: Journal of Nuclear Materials, to
be published (2011).
237
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Interface element) (Grain orientation) (Grain boundary) (Finite element method)
(Stress concentration)
Fig. 1 Finite element model for analysis. Fig. 2 Opening and shear stresses related to displacement.
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* Graduate School of Engineering, Osaka University, Japan Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
** JWRI, Osaka University, Japan Japan
238
Preliminary numerical research of microstructural fracture behavior in metal by using
interface element
Table 1 Stiffness of hexagonal crystal for a-Fe.
239
Transactions of JWRI, Vol.39 (2010), No. 2
4. Conclusion References
In order to demonstrate the microstructural fracture [1] H. Serizawa et. al: Trans. of JWRI, 33, 2, (2004), pp.181-185.
behavior in metal, the finite element method with an [2] N. Igata: Strength of Materials, Tokyo, BAIFUKAN Co., Ltd,
interface element was developed. The anisotropy in grain (1983), pp.11 (in Japanese).
[3] T. Tanaka et. al: Bulletin of the Japan Institute of Metals
and that at grain boundary were modeled by the ordinary Materia Japan, 58, 4, (1994), pp.382-389 (in Japanese).
finite element and the interface element, respectively. By [4] H. Nakashima et. Al: TETSU-TO-HAGANE, 86, 5, (2000),
using this method, the influence of anisotropy in grain and pp.357-362 (in Japanese).
at grain boundary on the microstructural fracture behavior
were examined successfully.
240
Transactions of JWRI, Vol.39 (2010), No. 2
241
Thermoreversible colloidal gelation for direct-assembly of nanoparticles
colloidal gel pushed out as a filament through a tapered direct writing technique. We used the PEO-PPO-PEO block
QR]]OH GLDPHWHU ' ȝP E\ D SQHXPDWLF V\VWHP 7KH co-polymer as a gelation agent. Using the present colloidal
filament was deposited on a plate or deposited layers while gel, the 3D wood-pile structure of alumina nanoparticles
the position of the nozzle was moving with a CAD system was directly formed.
control. The deposition was carried out in air, and the plate
for deposition was warmed at 60°C. The deposited colloidal
gel was dried with keeping the assembled shape.
The CAD program was designed to form the 3D-woodpile ȝP (a)
structures. The width between lattices was set 500ȝm.
(Sintering)
After drying, the assembled specimen was burned to
ȝP
remove polymers at 500°C for 5h and then sintered at
1450°C for 2h.
ȝP
3. Result and discussion
Figure 1 shows shear stress as a function of shear rate
for the prepared colloidal ink at 5 and 30°C. Viscosity of the
colloidal ink drastically increased with increase of
temperature. The shear stress at the low shear rate (0.1s-1)
(b)
was about 100Pa (at 30°C). It was applicable for the DIW.
Actually, the 3D wood-pile structure was successfully
formed without any defects.
104
103 30Υ
102
Shear stress (Pa)
ȝP
101
100 (c)
10-1
10-2 5Υ
10-3
10-1 100 101 102
ȝP
Shear rate (s-1)
Fig. 1 Shear stress as a function of shear rate for Fig. 2 Photographs of the assembled wood-pile structure
the prepared colloidal inks after sintering at 1450°C
(a) optical microscope (b),(c) SEM
Figure 2 shows the photographs of the specimen
assembled by the DIW. Fig.2(a) was taken by optical References
microscope. Fig.2(b) and Fig.2(c) were taken by scanning [1] J.A.Lewis and J.E.Smay, J.Stuecker, J.Cesarano : J. Am.
electron microscope (SEM). The wood-pile structure was Ceram. Soc., 89(2006), pp.3599-3609.
formed according to the CAD program. The filament which [2] M.Allahverdi, S.C.Danforth, M.Jafari and A.Safari : J. Euro.
was pushed out through the nozzle (diameter, D=1ȝP Ceram. Soc., 21(2001) , pp.1485-1490.
swelled a little because of release from the pressure at the [3] J.Cesarano, R.Segalman and P.Calvet : Ceram. Ind., (1998),
nozzle. After sintering, the diameter of the sintered filament pp.94-102.
was 125ȝP. The width between the lattices of the sintered [4] J.E.Smay, J.Cesarano and J.A.Lewis : Langmuir, 18(2002) ,
pp.5429-5437.
specimen was about 380ȝP. The shrinkage with the
[5] G.M.Lous, I.A.Cornejo, T.F.MacNulty, A.Safari and
sintering was about 20%. S.C.Danforth : J. Am. Ceram. Soc., 83, (2000), pp.124-128.
[6] S.L.Morissette, J.A.Lewis, P.G.Clem, J.Cesarano and
4. Conclusions D.B.Dimos : J. Am. Ceram. Soc., 84 (2001), pp.2462-2468.
In this study, we have newly developed [7] Qi Li and J.A.Lewis : Ad. Mater. , 15(2003),pp.1639-1643.
thermoreversible viscoelastic slurries which may provide
homogeneous colloidal gel states of nanoparticles for a
242
Transactions of JWRI, Vol.39 (2010), No. 2
†
Dispersion control of magnetic nanoparticles for functional fluids
YAMANAKA Shinya*, ABE Hiroya*, NAITO Makio*, UESHIMA Yuya**, NOMA Junichi**
243
Dispersion control of magnetic nanoparticles for functional fluids
4. Conclusions
We have found that the adequate addition of oleic acid
into silicone oil including the magnetic nanoparticles
dramatically reduced their apparent viscosity and was a key
to develop
colloidal
dispersion
system.
Additionally,
the
Fig. 2 Shear stress versus shear rate relationship with and resulting fluid clearly demonstrated a large MR effect.
without magnetic flux for 15 vol% MR fluids
244
Transactions of JWRI, Vol.39 (2010), No. 2
245
Molecular dynamics analysis of fcc nanowire under torsional loading
maintained at a constant 300 K by the velocity scaling face-centered-cubic (fcc), local hexagonal-close-packed
method. (hcp) or the other. In this specimen, it is the hcp atomic
structure that makes up the stacking faults. These hcp
3. Results and Discussion atoms are named as stacking-faults atoms and represented
atomic layer of the specimen with respect to time. The in red. Other atoms are named as other defects and
atomic layers are sequentially numbered along the z-axis. represented in blue. Fcc atoms are represented in gray.
The rotation angle is evenly distributed until approximately Elastic deformation occurs at 20 ps. The external force
30 ps. Then they are divided into two. causes equal strain along the twist axis. At 40 ps, multiple
Next, Fig. 3 shows the atomic configuration of a quarter stacking faults are observed, and at 60 ps, the deformations
of the Cu specimen at 20–80 ps. In order to analyze the are found to be concentrated in some of the stacking faults.
defect structure, common neighbor analysis (CNA) is used These concentrations give rise to localization of defect
to classify the structure of the Cu atoms into local structures. Finally, at 80 ps, the defect structures are almost
fully localized. The other stacking faults recover from their
deformations. As a result, we conclude that the given load
induces only localized defect structures.
Next, Fig. 4 and 5 show the change in the potential
energy per unit length, along the z-axis. It shows plots for
the entire system and for sets of three adjacent atomic
layers. Figure 4 shows the change in the potential energy
with time. Up to approximately 30 ps, the potential energy
increases quadratically with time. That is, the potential
energy increase quadratically with the macroscopic torsion,
and all atomic layers of the crystalstructure exhibitelastic
deformation. Then, up to approximately 120 ps, the 34–36
atomic layers show an increase in their twist angle, but the
other atomic layers show none. Hence, we can confirm that
macroscopically, the nanowire exhibits plastic
Fig. 4 Potential energy vs time. deformation; however, from the microscopic point of view,
the stacking fault elastic recovery is going. Next, Fig. 5
shows the relative twist angle for each set of three adjacent
atomic layers. The twist angle of the atomic interplane is
UHSUHVHQWHG E\ į UDGQP 7KH UHODWLRQ EHWZHHQ WKH WZLVW
angle and the potential energy for atomic layers is obtained.
As a result, we are able to define the microscopic
interplanar potentials.
4. Conclusions
We analyze fcc Cu nanowires under torsional loading
using molecular dynamics analysis. We investigate the
mechanism of localization by observing the potential
energy of each atomic layer and its twist angle in detail.
We obtain the following findings through this study:
(1) We demonstrate that macroscopic plastic deformation
Fig. 5 Potential energy vs torsion angle. occurred in the nanowires, but that from the microscopic
246
Transactions of JWRI, Vol.39 (2010), No. 2
point of view, the stacking fault elastic recovery is going. [3] T. Ogawa and A. Nakatani: International Symposium on
(2) We obtain the relation between the twist angle and the
Structures under Earthquake, Impact, and Blast Loading 2008
potential energy for each of the atomic layers. That helps
us define the microscopic interplanar potentials. (IB2008), (2008), pp. 47-52.
[2] O. Nguyen and M. Ortiz: Journal of the Mechanics and [5] Y. Mishin, M. J. Mehl, D. A. Papaconstantopoulos, A. F.
Physics of Solids, 50 (2002), pp. 1727-1741. Voter and J. D. Kress: Phys. Rev. B 63 (2001), p. 224106.
247
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Dental crown) (Alumina) (Stereolithography) (Flexural strength) (Sintering behavior)
248
Visualization of mechanical properties in alumina dental crowns fabricated by using laser scanning
stereolithography
5. Acknowledgments
This study was supported by Priority Assistance for the
Formation of Worldwide Renowned Centers of Research -
The Global COE Program (Project: Center of Excellence
for Advanced Structural and Functional Materials Design)
from the Ministry of Education, Culture, Sports, Science
and Technology (MEXT), Japan.
References
[1] M. Suwa, S. Kirihara and T. Sohmura: Ceramic Transactions,
219 (2009), pp. 331-336.
[2] M. Suwa, S. Kirihara and T. Sohmura: Proceedings of the
34th International Conference on Advanced Ceramics and
Composites, 31, 8(2009), pp.239-245.
Fig. 3 Cross sectional ceramic microstructures of alumina
flexural test samples with or without glass infiltrations. The
plate specimens were formed by stereolithography and heat
treatment processes. FS and RD mean the flexural strength
and relative density, respectively.
249
Transactions of JWRI, Vol.39 (2010), No. 2
, , ,
TAKENAKA Kosuke * ****, CHO Ken * ****, SETSUHARA Yuichi * ****,
, , ,
SHIRATANI Masaharu ** ****, SEKINE Makoto *** **** and HORI Masaru *** ****
KEY WORDS: (Plasma) (Polymer) (Combinatorial analysis) (Soft material) (Flexible electronics)
Antenna
plasmas with density gradient via localized power
180
200
250
Combinatorial analysis of plasma-polymer interactions for formation of inorganic-soft materials
hybrid structure
(b)
(a)
Intensity ratio
0.5
Ra (nm)
C-O/C-C
2 0.4
1.5 0.3 raw
1 0.2 88W
0.5 175W
0.1
0 350W
350
0
O=C-O/C-C
Intensity ratio
RSm (nm)
300 0.5
O=C-O/C-C
250 0.4
200 raw
150 0.3
88W
100 175W 0.2
50 350W 0.1
0
10
-2
10
-1 0
10 0
10-2 10-1 100
Iis (mA) x Time (min) Iis (mA) x T (min)
Fig. 3 Variation of surface roughness Ra and average Fig. 4 Variation in deconvoluted peak-area ratio of IC-O/IC-C
peak interval RSm of polymers on the product of and IO=C-O/IC-C as a function of the product of (ion
(ion saturation current Iis) x (plasma exposure saturation current Iis) x (plasma exposure time T) as a
time T) as a measure of the ion dose. measure of the ion dose onto the PET surface.
The XPS analyses were performed in AXIS-165x PET films and the PET surface located at x = 220 mm (I is =
spectrometer (SHIMADZU corp., JAPAN) with non- 0.43 mA) which was exposed to argon-oxygen plasma for 1
monochromatized MgKa radiation (photon energy of min. After exposure to the argon-oxygen plasma, the PET
1253.6 eV). surface showed enhancement in surface roughness (R a ) and
suppression of average peak interval (R Sm ) as can be seen in
3. Results and discussion AFM images. Figure 3 shows dependence of the surface
In order to investigate characteristics of argon/oxygen roughness (R a ) evaluated from the AFM images on ion
mixture plasmas, plasma parameters were measured for saturation current x exposure time corresponding to ion
argon/oxygen mixture plasmas. Considerable gradients of dose. The R a of the PET surface after plasma exposure
plasma density were obtained by localized power slightly increased with increasing ion dose, while the R Sm
deposition profiles (not shown here). Based on the results considerably decreased with increasing ion dose.
in terms of ion-saturation current density, analysis of Thr deconvoluted peak-area ratio of the C-O bond to
plasma-polymer interactions has been carried out in terms the C-C bond (I C-O /I C-C ) and O=C-O bond to the C-C bond
of etching characteristics and chemical bonding states. (I O=C-O /I C-C ) are evaluated and are summarized in Fig. 4 as
In order to examine the influence of plasma exposure a function of the product of the ion saturation current I is and
time on morphological change of polymers, the plasma- the plasma exposure time T as a measure of the ion dose
exposed surface was imaged with AFM. Figure 2(a) and onto the samples from the typical XPS C1s spectra of PET
(b) shows typical AFM images of the original surface of the films. These results indicate that the increase of the peak-
251
Transactions of JWRI, Vol.39 (2010), No. 2
area ratio I C-O /I C-C is considered to be due to C-O bond ion dose. The XPS results showed almost linear
formation either via the side-band scission of the O=C-O dependence with increasing ion dose. The data points
bond and/or via the incorporation of oxygen radicals and/or obtained from three independent batches of experiments
ions through chemical reactions on the surface. with 5 samples showed universal relations, indicating
efficiency of the combinatorial method for effective basic
4. Conclusions data acquisition.
Combinatorial analysis of plasma-polymer interactions
with combinatorial plasma-process analyzer has been References
carried out in terms of etching characteristics and chemical [1] International technology roadmap for semiconductors, 2007
bonding state of polymers. Surface roughness of the edition, Executive summary.
polymer slightly increased with increasing ion dose, while [2] Y. Setsuhara, T. Shoji, A. Ebe, S, Baba, N. Yamamoto, K.
the average peak interval considerably decreased with Takahashi, K. Ono, and S. Miyake, Surf. Coat. Technol.,
174-175 (2003), pp. 33-39.
252
Transactions of JWRI, Vol.39 (2010), No. 2
†
Observation of hole formation process in plasma arc drilling
KEY WORDS: (Plasma arc) (Drilling) (Hole formation) (Dross) (Molten metal)
1. Introduction
Recent thermal processes for drilling, such as laser
drilling and plasma arc drilling, are challenging the
traditional drilling methods. Laser drilling has high speed
and high precision for thin plates of almost all materials,
while plasma arc drilling has high speed and a high material
removal rate for thick plates of almost all metals.
The mechanisms of thermal drilling differ from those of
traditional drilling methods. Laser drilling as a
representative thermal process has been widely investigated.
Melt ejection during laser drilling of metals has been
observed and analyzed by high-speed photography,
showing a significant mechanism of material removal [1]. Fig. 1 Experimental setup.
When melt ejection occurs, spatter or dross readily forms at
the entrance and exit sides of the hole. The characteristics 3. Results and discussion
of spatter formation and its prevention have been Hole formation in plasma arc drilling is dependent on
investigated in detail [2, 3]. Furthermore, phenomena the removal of molten metal. Observations were carried out
accompanying hole formation, such as the flow velocity of at the upside and downside positions of the hole.
molten material and heat flow models, have also been Representative photographs are shown in Fig. 2. The
discussed [4, 5]. observation results show that the quantity of residual
The geometrical characteristics and drilling molten metal around the entrance hole increases with the
performance of the plasma arc drilling method were passage of processing time before the penetration time of
investigated in our previous studies [6, 7]. However, the 0.7 s. Subsequently, the quantity of residual molten metal
mechanism of hole formation during plasma arc drilling has remains almost unchanged and its temperature falls with the
not yet been fully elucidated. We therefore used a high- passage of processing time, as shown by the brightness of
speed video camera to observe and analyze the process of the white region. The process of ejection at the exit side
drilling under different conditions. The results obtained differs from that at the entrance side. Initially, the molten
confirm the mechanisms of the hole formation process, metal inside the hole is entirely blown away in the form of a
ejection of molten metal, and dross formation. cloud without residue immediately after penetration occurs.
This is because of the large internal pressure of the hole.
2. Experimental procedure Then, with the increase in the exit hole diameter, the
The setup for our plasma arc drilling experiments internal pressure decreases, leading to a reduction in the
consisted of a power supply device, gas supply device,
water-cooled tool, work-piece fixture, and time control
switch, as shown in Fig. 1. A high-speed video camera was
employed to observe the process of plasma arc drilling from
different positions (upside, horizontal, and downside). The
photographic conditions of the camera were fixed at a
frame rate of 2000 f/s and a shutter speed of 1/272,000 s. A
number of observational experiments were conducted by
varying the arc current (100, 120, 140 A) and torch height
(6, 8, 10 mm). Sequences of photographs with the passage
of time were obtained to analyze the drilling process.
253
Observation of hole formation process in plasma arc drilling
ejection of molten metal. As a consequence, molten metal Fig. 5 shows the representative appearances of a hole
begins to remain at some specific positions around the exit with dross and after dross removal at both the entrance and
side.In order to clearly observe the evolution of hole exit sides. Circular dross was formed around the entrance
formation, we conducted a series of drilling experiments at hole, while partial dross was produced around the exit hole.
the end surface. Representative photographs, which reveal Although the amount of residual dross at the entrance side
the processes of metal melting and hole formation, are is greater than that at the exit side, it is relatively easy to
shown in Fig. 3. The situation of metal removal and the remove. Comparing the appearances of both the entrance
development of the hole shape with the passage of and exit sides after the dross is removed, it can be seen that
processing time can be clearly observed. The hole shape the entrance side is cleaner while some indelible dross still
changes from a blind taper hole to a taper hole, then a remains at the exit side. This can be explained by the
cylinder hole, and finally a reverse taper hole. When the mechanism of ejection of the molten metal. As a
drilling processes under different conditions are compared, consequence of prolonged continuous processing, residual
the evolution of the hole is seen to be faster at the larger arc hot molten metal constantly remains around the exit side
current and lower torch height. Based on the above until the drilling is completed. The work-piece is partly
observations, the mechanisms of hole formation and melted and fuses with the residual molten metal. The
ejection of molten metal are simulated in Fig. 4. The combination makes dross removal difficult. At the entrance
plasma ejected from the nozzle heats the work-piece to side, however, the work-piece is not fused with the residual
melting point or boiling point. The molten metal is ejected molten metal, with the result that the dross and the work-
from the entrance side by the high-pressure gas (a). piece are readily separated.
However, because of its viscosity, the molten metal is not
entirely blown away and some remains around the entrance
side of the hole before penetration occurs (b). After
penetration, almost all of the molten metal is ejected from
the exit side (c), and the quantity of molten metal at the
entrance side thus remains unchanged. With the passage of
processing time, the exit hole diameter continues to enlarge
(d), with a cylinder hole (e) and then a reverse taper hole (f)
being formed. Meanwhile, some of the molten metal
remains around the exit hole.
4. Conclusion
Fig. 4 Mechanisms of hole formation and ejection of The process of hole formation during plasma arc
molten metal. drilling was observed using a high-speed video camera. The
254
Transactions of JWRI, Vol.39 (2010), No. 2
evolution of the hole from a blind taper hole to a taper hole, [2] D. K. Y. Low, L. Li, A. G. Corfe: J. Mater. Process. Tech.,
cylinder hole, and reverse taper hole was clearly observed 118 (2001), pp. 179-186.
under different conditions. The mechanisms of hole [3] D. K. Y. Low, L. Li, P. J. Byrd: J. Mater. Process. Tech., 139
formation with the passage of processing time and ejection (2003), pp. 71-76.
[4] C. F. Cheng, Y. C. Tsui, T. W. Clyne: Acta Metall. Mater. 46
of molten metal at the entrance and exit sides were clarified
(1998), p. 4273.
by analysis of the process phenomena. An understanding of [5] A. Luft, U. Franz, A. Emsermann, J. Kaspar: Appl. Phys. A,
these mechanisms will be useful to improve hole quality 63 (1996), p. 93.
and reduce dross around the hole. [6] K. Kusumoto, H. Sun: Nat’l. Meeting of JWS, 82 (2008-4), pp.
114-115.
References [7] H. Sun, K. Kusumoto: Nat’l. Meeting of JWS, 83 (2008-9), pp.
[1] K. T. Voisey, S. S. Kudesia, W. S. O. Rodden, D. P. Hand, J. 74-75.
D. C. Jones, T. W. Clyne: Mater Sci Eng, A356 (2003), pp.
414-424.
255
Transactions of JWRI, Vol.39 (2010), No. 2
†
Simulations and visualizations of grid erosion in ion engines
Prediction of ion engine lifetime
NAKANO Masakatsu *
KEY WORDS: (Prediction) (Numerical simulation) (Lifetime) (Grid erosion) (Sputtering) (Ion engine)
256
Simulations and visualizations of grid erosion in ion engines
0
Fig. 5 (b) Decel grid after 20,000 h of operation for I a = -350 V.
-50 accumulated operating time reaches the specified value. A
detailed description of the simulation is given in Ref. 1.
-100
-300 [V]
-150
3. Results and discussion
-350 [V] Parametric calculations for the P20 ion engine grid were
-200 -400 [V] performed by varying the accel grid voltage. Calculations
-450 [V] were performed for high and low ion beam currents and a
-250 fixed total acceleration voltage. Detailed grid and operation
0 10000 20000 30000 40000 parameters are given in Ref. 2.
Operation time [hrs] Figure 3(a) shows the grid surfaces and erosion
Fig. 4 Minimum potential on axis. intensities for grid holes in the high ion beam current
region at the beginning of life (BOL) and after 20,000 h of
charge-exchange and elastic collisions are evaluated from operation. The downstream edge of the accel grid hole is
the ion beam current, collision cross section, and local severely eroded, which may lead to electron backstreaming.
neutral density. Neutral densities are obtained prior to Fig. 3(b) represents the potential distribution and ion beam
performing the potential and beam trajectory calculations trajectories. From the figure, the downstream edge erosion
using the free molecular flow approximation. Sputtering of of the accel grid is explained by the impact of cross-over
the grid material is incorporated using experimental data in ions.
the form of tables of values for incident energies and Figure 4 shows the variation of the minimum
incident angles. Since redeposition of the sputtered grid electrostatic potential on axis as a function of the
materials is a very complex process that involves accumulated operating time for I a = -300 to 450V. If
interactions between the incident sputtered particles and the electron backstreaming is assumed to occur when the
grid surface atoms, it is modeled by a simplified approach minimum axial potential exceeds -T e [V] (T e : electron
using a sticking factor, in which the redeposition flux is temperature), the lifetimes are predicted to be about 14,000
given by the product of the incoming sputtered particles h for I a = -300 V, 32,000 h for I a = -350 V, and over
and the sticking factor.
40,000 h for I a = -400 and -450 V.
The grid surface profile is updated from the rate of
Figure 5 shows the grid surfaces at the BOL and at
change of the grid surface obtained by the sputtering and
after 20,000 h of operation for grid holes in the low ion
redeposition calculations. This update is repeated until the
beam current region. Since ion beams tend to diverge at
257
Transactions of JWRI, Vol.39 (2010), No. 2
lower beam currents, the inner surfaces of the accel and (2) Two major failure modes (electron backstreaming and
decel grids are significantly eroded and the decel grid is structural failure) of the P20 ion engine were
close to structural failure due to six-point star-shaped successfully analyzed.
erosion. (3) Using numerical simulations accelerates the lifetime
Based on these simulations, it is concluded that the end verification process by two orders of magnitude or more.
of life of the grid holes in the high beam current region is
caused by electron backstreaming, whereas the end of life Acknowledgement
of the grid holes in the low beam current region is caused This study was performed under a corporative research
by structural failure of the decel grid in the P20 grid system. program with JAXA. The author thanks Prof. Nishiyama
The total calculation time was about one week using a for providing the P20 ion engine data.
PC (Intel Core i7 965), which is two orders of magnitude
smaller than actual lifetime tests. References
[1] M. Nakano: Vacuum 83 (2008), pp. 82-85.
4. Conclusions [2] K. Nishiyama, Y. Toyoda, S. Hosoda, Y. Shimizu, and H.
The following conclusions were obtained: Kuninaka: Journal of Plasma and Fusion Research SERIES, 8
(2009), pp.1590-1594.
(1) Numerical simulations can be employed to determine
the lifetimes of ion engines.
258
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Materials Joining and Welding) (Laser processes) (Plasma processes) (Friction stir welding)
(Nanoparticles)
1. Research Activities of JWRI, Osaka University 3. Research Division of Materials Joining Mechanism
Since the foundation originally as “Welding Research This division has devoted research activities for
Laboratory” in 1969 at Osaka University, the Joining and investigation of the physical and chemical properties of
Welding Research Institute (JWRI) has devoted research high performance joints and composites produced by
activities to better understanding of fundamental welding and joining, and the control of the joint
mechanisms involved in joining and welding science and microstructure by highly-controlled material processing.
development of industrial technologies in wide scales of The results of these investigations have been applied to
structures ranging from huge constructions including the designing and development of optimum materials for
nuclear power plants, ships and tall buildings to micro and joining and welding. Investigation is also made of the
nano-scales including microelectronic devices and modification of the joint microstructure by well-controlled
nanoparticles, as key technologies for manufacturing materials processing using laser and thermal energies.
functional structures.
With a variety of high-energy-density beams (arc 4. Research Division of Functional Assessment
discharge, lasers and electron beams) as well as other types The structures studied in this division are not only
of heat source including frictional heat, JWRI has carried large-size welded structures including ships, bridges, and
out investigations on welding and joining, creation of nuclear reactors, but also microelectronic and nano
advanced materials and composites, highly functional structures, which have specific structures and functions, and
surface modifications, and development of novel frontier structures used in space and in the deep sea. The
technologies including space welding. researches progress under the balance between social
Since the restructuring of the JWRI in 1996, the JWRI usefulness and the conservaton of environment and
has consisted of three divisions (the research division of resources.
Materials Processing System, the research division of
Materials Joining Mechanism, and the research division of 5. Smart Processing Research Center
Functional Assessment) and the Smart Processing Research The Smart Processing Research Center was established
Center. in 2003 by consolidating two centers, i.e., the Research
Center for Ultra High Energy Density Heat Source, and the
2. Research Division of Materials Processing System Research Center for Materials Recycling and Integration, to
This division is oriented to provide scientific bases for develop advanced processes for nano and micro-structured
generation, conversion, transfer of processing energy materials, thereby contributing to the establishment of next
sources and interaction with materials for well-controlled
multi-functional processes and development of high-quality
energy sources suitable for various materials processing.
In particular, the division has made great contributions
to development of various advanced materials processing
systems with concentrated and dispersed energy sources,
which are expected to have wide application fields in the
future. Major emphasis is placed on the developments of
advanced monitoring and estimating techniques for material
processing and the systematization of the functional
materials processing technology through interpretation of
phenomena relating to the high energy involved in the
processing.
Fig. 1. Cooperation of the six Research Institutes.
259
Materials joining technologies and interface science for integration of novel structured metallic and inorganic
materials
generation manufacturing that can encourage industries to 7. Global Collaborative Research Center for Computational
minimize any negative impact on the environment. Welding Science (CCWS)
To establish the advanced science and technology in the The JWRI proposed the concept of computational
global research field of welding and joining processes, welding science as early as the 1970’s. It is one of the
JWRI is making continuous and significant efforts to obtain world pioneers in this field and has organized an
instructive and informative R&D results by intelligent international symposium on “Theoretical Prediction in
innovation. Joining and Welding” in 1996. In further development of
computational welding science, JWRI conducted a project
6. Development Base of Advanced Materials on “Development of Highly Efficient and Reliable Welding
Development and Integration of Novel Structured Technology” which has been supported by the New Energy
Metallic and Inorganic Materials and Industrial Technology Development Organization
The project, Development Base of Advanced Materials (NEDO). One of the main objectives of this project is to
Development and Integration of Novel Structured Metallic establish a framework of the computational welding science
and Inorganic Materials (Fig. 1), started in 2010 as inter- which covers the entire aspects of welding, i.e. Welding
university cooperative research project (Joining and Process, Welding Metallurgy and Welding Mechanics. This
Welding Research Institute, Osaka Univ., Institute for framework has been further developed to meet various
Materials Research, Tohoku Univ., Materials and demands from the industry.
Structures Laboratory, Tokyo Institute of Tech., Eco Topia
Science Institute, Nagoya Univ., Institute for Nanoscience 8. Research Activities for the Project “Advanced
and Nanotechnology, Waseda Univ., Institute of Materials Development and Integration of Novel
Biomaterials and Bioengineering, Tokyo Medical and Structured Metallic and Inorganic Materials”
Dental Univ.). This development base promotes the joint In the present project, research and development
research for development of new functional materials by activities are underway in the following fields of
integration of novel structured metallic and inorganic investigations; 1) Development of Materials for
materials for applications in the specific fields including Environment and Energy System, 2) Development of
environment&energy, electronics, and biomedical materials Electronics Materials, 3) Development of Bio-Medical
through the inter-university cooperative researches. Materials, 4) Development of Novel Structured Advanced
Materials and Integration Technologies.
Fig. 3. Soldering technologies. Fig. 4. Friction stir welding of bulk metallic glasses.
260
Transactions of JWRI, Vol.39 (2010), No. 2
In the research field for Environment and Energy technologies. For the development of Bio-Medical
System, JWRI is to contribute to develop Environment Materials, JWRI is to make approaches via nano-particle
and Energy Systems with metallic glasses and advanced technologies and bio-ceramics coating with advanced
ceramics via nano-assembly technologies (nano-particles, surface modification technologies.
slurry and casting, Fig. 2), micro-assembly technologies, Furthermore, in the field of Development of Novel
dissimilar materials joining technologies based on laser Structured Advanced Materials and Integration
welding and nano-surface modification with advanced Technologies, a variety of research activities are to be
plasma processing. carried out for the establishment of scientific and
For development of Electronics Materials, the JWRI is technological bases of joining of metallic glasses and
to contribute to develop device fabrication technologies via advanced ceramics materials. (Fig. 4)
soldering technologies (Fig. 3) and plasma processing
261
Transactions of JWRI, Vol.39 (2010), No. 2
FUKUHARA Mikio*
KEY WORDS: (Cooperative research) (6 Universities) (Environment and energy materials) (Electronic
materials) (Bio- and medical materials) (Tohoku university’s activities)
262
Cooperative researches in Tohoku University
Fig.1 The summary figure of the 2nd project, “Advanced Materialsᴾ Development
and Integration of Novel Structured Metallicᴾ and Inorganic Materials” derived
from the 1st projectᴾ “Research and Development Project on Advanced Metallicᴾ
Glasses, Inorganic Materials and Joining Technology”.
263
Transactions of JWRI, Vol.39 (2010), No. 2
OKADA Kiyoshi *
KEY WORDS: (Advanced inorganic materials) (Novel structures) (Environment) (Energy) (Electronics)
(Biomaterials)(MSL) (Tokyo Tech)
264
Approach for the project on advanced materials development and integration of novel structured metallic and
inorganic materials by materials and structures laboratory of Tokyo Institute of Technology
3. Research topics in the MSL institutes are very important to enhance our research works
The main topics planned in this special project at the toward the goal.
MSL are listed in Table 1.
Environment and energy materials: Since the main topics in References
this field are focused on the materials related to fuel cells [1]E.Yasuda (eds): Ceramics and Building Materials, Yokohama,
organized by the ETSI, we will investigate a material for Materials and Structures Laboratory, (2004).
separators using metallic glasses [3]. On the topics related [2] A.Inoue (eds): Research Result Report of Metallic Glasses,
Sendai, Tohoku University, (2010).
to the environment, a collaborative work on oxidation
[3] K.Katsumata, T.Wada, A.Inoue, M.Sakai, A.Nakajima,
catalyst using mayenite type compounds [4] will be A.Fujishima, N.Matsushita and K.Okada: Abstracts
performed with the ETSI. Book STAC-4, Yokohama, (2010), pp.
Electronics materials: Main part of the work in this field [4]J.Li, K.Hayashi, M.Hirano and H.Hosono: Solid State Ionics,
is preparation and evaluation of photonic, electronic and/or 180 (2009), pp. 1113-1117.
spintronic devices with some novel nanostructures by [5]T.Kamiya, K.Nomura and H.Hosono: Phys. Status Sol. A, 206
combinations of inorganic materials, semiconductors and/or (2009), pp. 860-867.
metals [5-8]. Since these devices consist of multi-thin film [6]T.Taniyama, K.Akasaka and M.Itoh: J. Appl. Phys., 105
nanotextures, strong collaboration with the INN group is (2009), pp. 07D901-1-3.
[7]R.Takahashi, M.Katayama, O.Dahl, J.K.Grepstad,
essentially important.
Y.Matsumoto and T.Tybell: Appl. Phys. Lett., 94 (2009), pp.
Bio and medical materials: Preparation and biomedical 232901.
evaluation of Ti based bulk metallic glasses (BMGs) having [8]I.M.Vishik, E.A.Nowadnick, W.S.Lee, Z.X.Shen, B.Moritz,
nanostructured bioactive surfaces is the main topic and this T.P.Devereaux, K.Tanaka and T.Sasagawa: Nature Phys., 5
will be performed by collaborative team of the MSL, IMR, (2009), pp. 718.
IBB and JWRI. The MSL will mainly take a part of [9]N.Sugiyama, H.Xu, T.Onoki, Y.Hoshikawa, T.Watanabe,
preparation of nanostructured bioactive surfaces on the Ti- N.Matsushita, X.Wang, F.X.Qin, M.Fukuhara, M. Tsukamoto,
BMGs using hydrothermal and/or electro-hydrothermal N.Abe, Y.Komizo, A.Inoue and M. Yoshimura: Acta
techniques [9,10]. Evaluation of bioactivity of oxide thin Biomater., 5 (2009), pp. 1367-1373.
films controlling their atomic precisions [11] is a new [10] T.Onoki, T.Higashi, X.Wang, S.Zhu, N.Sugiyama, Y.
Hoshikawa, M.Akao, N.Matsushita, A.Nakahira, E. Yasuda,
challenging work in this field. M.Yoshimura and A.Inoue: Mater. Trans., 50 (2009), pp.
Novel structured materials: Relatively wide topics are 1308-1312.
proposed in this field but main topics will be focused on [11]T.Susaki, S.Kumada, T.Katase, K.Matsuzaki, M. Miyakawa
control of interfaces for developing novel structured and H.Hosono: Appl. Phys. Exp., 2 (2009), pp. 091403.
materials [12-14]. The topics related to joining of different [12]Y.Shinoda, Y.Yanagisawa, T.Akatsu, F.Wakai and H.Fujii:
materials will be performed collaboratively with the JWRI. Mater. Trans., 50 (2009), pp. 1250-1254.
[13]N.Kawai, T.Atoh, K.G.Nakamura, K.Kondo, S.Ito, K. Yubuta
4. Conclusions and M.Kikuchi: J. Appl. Phys., 106 (2009), pp. 23525.
In the present project, clear outcomes developed to [14]T.Yoshida, Y.Moriya, T.Tojo, H.Kawaji, T.Atake and
Y.Kuroiwa: J. Therm. Anal. Calor., 95 (2009), pp. 675-683.
actual materials contributing to our society are strongly
requested. Strong collaborations with the six member
265
Transactions of JWRI, Vol.39 (2010), No. 2
YOGO Toshinobu *
Exo.
(b)
2. Synthesis of Hybrid Membranes 80
Starting compounds for membranes are styryl-
Weight / %
266
Synthesis and properties of hybrid-type polymer membranes for fuel cells
The results confirm that the hybrid membranes possess and Si-O linkage were successfully synthesized via the
sufficient thermal stability up to 200ºC. It is worth noting copolymerization and sol-gel reaction. A phosphonic acid
that the exothermic peak temperatures vary according to the group was immobilized to polymer chain by radical
functionality of the membrane. The DTA curve for polymerization of styryl-substituted alkoxysilane and a
DMMSMS(M)/EPA=1/2 membrane exhibits a peak at derivative of acrylic phosphonic acid (EPA) via a stable C-
328ºC, whereas the curve for MDMSMS(D)/EPA=1/2 P bond. The membranes exhibited good thermal stability,
membrane shows a peak at 348ºC. MDMSMS(D)/EPA which depended on the degree of inorganic chain in the
hybrid membrane has a linear siloxane linkage and a higher hybrid membrane. The proton conductivity depended on
thermal stability than that of DMMSMS(M)/EPA above the temperature and relative humidity. The membranes
280ºC. This result reveals that siloxane chain formed in synthesized from monomethoxysilane derivative were
the membrane improves the thermal stability of the hybrid found to show higher conductivities than those from the
membrane. dimethoxysilane derivative. The membrane exhibited
Figure 2 shows the proton conductivities under good conductivities up to 140ºC at various humidities.
different relative humidities for the hybrid membranes of The hybrid membranes immobilizing phosphonic acid
DMMSMS(M)/EPA=1/6 and MDMSMS(D)/EPA=1/6. The groups have a potential for a PEFC application operated at
conductivities of Nafion 112 measured under the same intermediate temperatures.
humidified conditions are also shown in Fig. 2. Obviously,
R
the conductivity strongly depended on the relative humidity Temperature / C
and decreased with decreasing relative humidity. The -1140130120 100 80 60 40
proton conduction for hydrated membranes generally 10
occurs via a proton hopping mechanism (Grotthuss
mechanism) and a diffusion mechanism (vehicle (a)
mechanism) [1]. It should be noted that the activation -2
10
energy of the proton conduction is dependent on the relative
-1
Conductivity / S cm
humidity and increases with decreasing the relative (b)
humidity. The activation energies for
DMMSMS(M)/EPA=1/6 and MDMSMS(D)/EPA=1/6 -3
10
membranes are 0.14 and 0.26 eV at fully hydrated state,
0.40 and 0.48 eV at about 65 %RH and 0.75 and 1.05 eV at
about 20 %RH, respectively. At the fully hydrated state, (c)
-4
the activation energy for proton conduction in the hybrid 10
membrane is comparable to that of Nafion [2]. This result (d)
suggests that water molecules absorbed in the membrane (e)
contribute to the proton conduction like usual water- -5
(f)
cooperative proton conductive membranes by both 10
Grotthuss and vehicle mechanisms in the fully hydrated 2.5 2.6 2.7 2.8 2.9 3.0 3.1 3.3
-1
state. The increase in activation energy with decreasing 103(1/T) / K
relative humidity is reported for Nafion [2]. Similarly, the Fig. 2 Temperature dependencies of proton conductivities
increase in activation energy for the current membrane for the hybrid membranes of
derives from the decrease in water content used for proton DMMSMS(M)/EPA=1/6 (solid symbols) and
conduction. The conductivities of hybrid membranes were MDMSMS(D)/EPA=1/6 (open symbols) under
considerably low compared to that of Nafion 112 in the different relative humidities: (a) and (b) 100 %RH,
range from 40 to 120ºC (Fig. 2). However, the (c) and (d) from 63.5 to 67.7 %RH humidified with
conductivity of Naffion 112 drastically decreased above NaCl solution, (e) and (f) from 19.3to 20.8 %RH,
120ºC. In contrast, the conductivity of hybrid membranes humidified with MgCl 2 solution.The conductivities
increases with temperature up to 140ºC under any of Nafion 112 are also plotted for comparison; (+)
humidified conditions. Also, no decrease in conductivity humidified with NaCl solution, (×) humidified with
by water evaporation was observed up to 140ºC even at low MgCl 2 solution.
humidified conditions. The results indicate that the
membrane is superior in water retention to Nafion above References
120ºC. [1] W.H.J. Hogarth, J.C.D. da Costa, G.Q. Lu: J. Power Sources,
142 (2005), pp.223-237.
4. Conclusions [2] Z. Siroma, R. Kakitsubo, N. Fujiwara, T. Ioroi, S. Yamazaki,
Proton conducting hybrid membranes consisting of K. Yasuda: J. Power Sources, 189 (2009), pp.994-998.
organic polymer chains containing phosphonic acid groups
267
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Laser Welding) (Dissimilar materials joining) (Steel) (Aluminum alloy) (Plastic) (TEM image)
268
Evolution of laser welding to dissimilar materials joining
weld bead on both surfaces of A5052 and low carbon steel, that the carbon steel and the plastic are tightly bonded at the
and showed good appearances as shown in Fig. 3. atomic or molecular size level.
On the other hand, as for the boundary between A5052
and PET in the obtained joint, the observation image is
shown in Fig. 5. The bright upper part is PET and the
dark lower area is A5052 aluminum alloy. The TEM
image also demonstrates that the aluminum alloy and the
plastic are tightly bonded at the atomic or molecular size
level, which was similar to the above-mentioned LAMP
joints. Moreover, it was found that the oxide film was
composed mostly of aluminum and magnesium by energy
dispersive X-ray spectroscopy (EDS) analysis as indicted in
Fig. 6. A carbon peak was measured in the EDS analysis,
which means that the oxide film of A5052 and the polymer
molecule of PET are close on the atomic or molecular size
level.
㻝㻜㻜 㻸㻳㻟㻜㻜㻞㻟㻌㻧㻌㻞㻟㻚㻟㻤㻜㻌㼚㼙
㻡㻜
㻻
㻠㻜
㻹㼓
㻟㻜
㻹㼛
㻞㻜
㻹㼛
㻝㻜
㻜
㻜㻚㻜㻜 㻝㻚㻜㻜 㻞㻚㻜㻜 㻟㻚㻜㻜 㻠㻚㻜㻜 㻡㻚㻜㻜 㻢㻚㻜㻜 㻣㻚㻜㻜 㻤㻚㻜㻜 㻥㻚㻜㻜 㻝㻜㻚㻜㻜
㼗㼑㼂
Acknowledgments
This work was conducted as part of the Industrial
Technology Research Grant Program “Development of
high-quality direct laser joining of metal and plastic by
laser adaptive control” supported by New Energy and
Industrial Technology Development Organization (NEDO).
The authors would like to acknowledge whom it may
concern.
References
[1] S. Katayama and Y. Kawahito, Scripta Materialia, 59-12
(2008), pp.1247- 1250.
[2] S.Katayama; Y. Kawahito; A. Tange; S. Kubota, Online Proc.
Fig. 5 TEM image of boundary between aluminum alloy and of LAMP 2006, Kyoto, (2006) #06-7.
PET in dissimilar metal weld.
269
Transactions of JWRI, Vol.39 (2010), No. 2
†
Introduction to vector materials science and bioengineering
YAMASHITA Kimihiro *
270
Introduction to vector materials science and bioengineering
271
Transactions of JWRI, Vol.39 (2010), No. 2
[2] K. Yamashita, N. Oikawa and T. Umegaki: Chem. Mater., 8 [5] M. Ueshima, S. Nakamura, M. Ohgaki and K. Yamashita:
(1996), pp.2697-2700. Solid State Ionics, 151 (2002), pp.29-34.
[3] M. Ohgaki, S. Nakamura, T. Okura and K. Yamashita: J. [6] Y. Toyama, M. Ohgaki, S. Nakamura, K. Katayama and K.
Ceram. Soc. Jpn., 108 (2000), pp.1037-1040. Yamashita: Solid State Ionics, 151 (2002), pp.159-163.
[4] H. Takeda, Y. Seki, S. Nakamura and K. Yamashita: J. Mater. [7] A. Obata, S. Nakamura, Y. Moriyoshi and K. Yamashita: J.
Chem., 12 (2002), pp.2490-2495. Biomed. Mater. Res., 67A (2003), pp.413-420.
272
Transactions of JWRI, Vol.39 (2010), No. 2
1. Introduction
Table 1 Bath composition for Pt
Metal nano-particles have been known to possess electrodeposition.
various properties which cannot be obtained in the bulk
Chemicals Concentration ᴾ ᴾ ᴾ ᴾ
phase and are expected to be applied to various fields such
as electronics and biotechnology, etc [1-4]. The nano H 2 (PtCl 6 ) · 6H 2 O ᴾ 10 g dm-3
particles are prepared through physical or chemical
approaches. The chemical method has advantages for PEG (M. W. 200- ᴾ 0-100 mM
fabricating the nano particles with precisely controlled nano 20000)
structure. Among the chemical methods, an electrochemical
process is utilized in which the nano particles are prepared Figure 1 shows the UV-NIL process for the nano-
by reduction at a cathode substrate [5]. In this case, the patterning of the cathode surface patterns. First, UV resin is
metal occasionally forms dendritic like growth, which spin-coated on a Si substrate with a seed layer (1). A
results in non-uniformity of the particles. In order to solve stamper with nano patterns is pressed to the substrate and
this problem, in this study, we attempt to optimize the UV light is irradiated from above (2, 3, 4). Then, the
surface nanostructure of the cathode. In order to prevent stamper is removed from the sample (5). Subsequently, the
coagulation of the metal particles and achieve numerous residual films at the hole are removed using O 2 RIE (6).
nucleation, we applied the nano-dot-patterning to the This substrate with a nano-patterned surface is used as a
electrode surface using Pt electrodeposition and UV-NIL cathode electrode(7). Figure 2 shows the schematic
processes[6]. Furthermore, we investigated the Pt illustration of the Pt electrodeposition. A Pt film is
deposition conditions, mainly focusing upon the effect of deposited on the Si substrate with nano patterns using a
rotating disk electrode system. The 100 nm thick Pt films
additive such as polyethylene glycol(PEG) in order to
were deposited using HZ-5000 rotating electrode system
obtain high surface activity.
(HOKUTO DENKO) with a SCE as reference electrode.
2. Experimental
A Ni and a Ti seed layers were deposited onto Si(111)
wafer using sputtering. The thickness of the Ni and the Ti
layers were 100 nm and 10 nm, respectively. The substrates
with the seed layers were cleaned with 14 wt% HCl
aqueous solution for 20 s followed by rinsing with ultra
pure water, then in order to generate uniform nucleation the
specimens were immersed into a 0.2 mM Pd aqueous
solution for 300 s followed by rinsing with ultra pure water.
After these pretreatments, the Pt was electrodeposited using
the bath shown in Table 1. The surface morphology was
observed using scanning electron microscopy with field
emission source (FE-SEM, S-4800, Hitachi) and atomic
force microscopy (AFM, Dimension 3100, Veeco
Instruments). The composition was analyzed using a glow
Fig. 1 UV-NIL process.
discharge optical emission spectrometer (GDOES, JY-
5000RF, HORIBA).
273
Preparation of metal nano particles using electrochemical deposition
Pt nano patterned electrodes
molecular weight (M. W.) of the PEG. These results for the
-0.2
behavior of PEG showed the same trends to those observed
Electrodeposited Pt film
for Cu electrodeposition for via-filling [7]. deposited without additives
-0.3
Figure 4 shows representative AFM images of the Electrodeposited Pt film
electrodeposited Pt films. As a reference, the image for deposited with PEG6000, 10
-0.4 mM
sputtered Pt film is also shown. From the results, it is Sputtered Pt film
confirmed that the electrodeposited Pt films have rougher
-0.5
surface than that of the sputtered film. Moreover, the -1.8 -1.6 -1.4 -1.2 -1 -0.8 -0.6 -0.4 -0.2 0
roughness became larger by the addition of PEG to the bath.
On the other hand, as is seen in Fig. 4, when the PEG with Potential / V vs. SCE
larger M. W. such as 20000, is applied, the surface Fig. 5 Cathodic polarization measurement of the Pt films in
becomes smoother. 0.5 M aqueous sulfuric acid.
0.02
0
Current Density / Acm -2
-0.02
-0.04
-0.06
No additive
-0.08
PEG6000, 10 mM 200 nm 200 nm
-0.1
PEG20000, 10 mM ᴾ (a) Pt nano patterned (b) Cu nano-particles formed
-0.12
on the Pt nano patterned
-2 -1 0
Potential / V䚷䡒䡏. SCE
Fig. 6 SEM images for electrodeposited Pt nano patterned
Fig. 3 Cathodic polarization measurement of the bathᴾ for electrode using UV-NIL(a) and Cu nano-particles formed
the Pt electrodeposition. on the Pt nano patterned(b).
274
Transactions of JWRI, Vol.39 (2010), No. 2
275
Transactions of JWRI, Vol.39 (2010), No. 2
ITOH Takashi *
KEY WORDS: (Eco-friendly) (Higher manganese silicide) (Magnesium silicide) (Synthesis) (Mechanical
grinding) (Pulse discharge sintering) (Liquid-solid phase reaction) (Thermoelectric properties)
276
Fabrication and thermoelectric properties of eco-friendly silicides for thermoelectric power generation using waste
heat
6.355 0.2
0.1
6.350
Lattice parameter (Å)
0.0
300 400 500 600 700 800 900
6.345
Temperature (K)
277
Transactions of JWRI, Vol.39 (2010), No. 2
substituted into parts of Mg sites in Mg2Si. The maximum Metastable and Nanocrystalline Materials, 8 (2000), pp.918-
ZT of 0.61 was achieved at 883K by optimizing Al content 923.
(y=0.15 at.%) in Mg2Si. [3] T. Itoh and M. Yamada: J. Electronic Materials, 38 (2009),
pp.925-929.
References [4] M. Riffel and J. Schilz: Proc. 16th Int. Conf. on
[1] S. Okada, T. Shishido, M. Ogawa, F. Matsukawa, Y. Thermoelectrics, Dresden, IEEE, (1997), p.283.
Ishizawa,K. Nakajima, T. Fukuda and T. Sundstrand: J. Cryst. [5]T. Kajikawa, K. Shida, M. Ohmori and T. Hirai: Proc. 16th Int.
Growth, 229 (2001), pp.532-536. Conf. on Thermoelectrics, Dresden, IEEE, (1997), p.275.
[2] M. Umemoto, Z.G. Liu, R. Omatsuzawa and K. Tsuchiya: J. [6] J. Tani and H. Kido: Intermetallics, 15 (2007), pp.1202-1207.
278
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Solid electrolytes) (Zirconium oxide) (Ordered porous structures) (Micro stereolithography)
2. Experimental Procedure
The zirconia lattice model with the coordination Fig. 2 A schematically illustrated micro stereolithography ᴾ
number 4 was created by the computer graphic application system. Acrylic resin lattices including zirconia nano-sized ᴾ
(Toyota Keram, Think Design) as shown in Fig. 1. The particles were processed through solidified layers stacking.
†ᴾ Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
*ᴾ JWRI, Osaka University, Ibaraki, Japan Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
Japan
279
Fabrication of zirconium oxide solid electrolytes with ordered porous structures by using micro stereolithography
4. Conclusions
We have fabricated three dimensional solid electrolytes
of zirconia lattices successfully by using stereolithography.
These ceramic lattices with coordination number 4 could
Fig. 4 The sintered zirconium oxide lattices with the ordered
realize ordered porous structures. The fabricated ceramic
porous structures. The lattice coordination number is 4. The
structures with high porosities and wide surface areas are
opened air paths were arranged periodically.
expected to be applied to novel electrodes in solid fuel cells.
References
[1] C. Kai and L. Fai: Rapid Prototyping, Singapore John Wiley
5. Acknowledgments and Sons, (1997).
This study was supported by Priority Assistance for the [2] A. Pique, A. Holmes and D. Dimos: Materials Research
Formation of Worldwide Renowned Centers of Research - Society Symposium Proceedings, (2003), pp.758-764.
The Global COE Program (Project: Center of Excellence [3] S. Kanehira, S. Kirihara and Y. Miyamoto, Journal of the
for Advanced Structural and Functional Materials Design) American Ceramic Society, 88 (2005), pp.461-1464.
[4] S. Kirihara and Y. Miyamoto, The International Journal of
from the Ministry of Education, Culture, Sports, Science
Applied Ceramic Technology, 6 (2009), pp.41-44.
and Technology (MEXT), Japan.
280
Transactions of JWRI, Vol.39 (2010), No. 2
SUZUKI Kenzi *
80 80
Toluene Conversion(%)
60 60
Intensity(a.u.)
Ni/Mayenite
40 40
Ni/Mayenite(5%)
20 20 Ni/Mayenite(3%)
Ni/Mayenite(1%)
Mayenite
0 0
550 600 650 700 750 800 850
Mayenite 550 600 650Temperature(
700 0750 800 850
C)
Temperature (䁸C)
900 1000 1100
-1
Raman Shift(cm )
Ÿ1L0D\HQLWHƔ1L0D\HQLWH
Ŷ1L0D\HQLWHż0D\HQLWH
Fig. 1 Raman spectra of Mayenite and Ni/Mayenite Fig. 2 Toluene conversion vs. Reaction temperature for
before reaction. Mayenite and Ni/Mayenite.
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* EcoTopia Science Institute, Nagoya University, Nagoya, Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
Japan Japan
281
Application to catalyst of mayenite consisting of ubiquitous elements
100
10ÛC/min in air. The weight loss due to the combustion of
Toluene conversion (%)
70 70
H2 yield, CO, CO2 selectity
80
282
Transactions of JWRI, Vol.39 (2010), No. 2
283
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Carbon Nanotubes) (Pd Nanoparticles) (Arrangement) (SDS) (Supramolecular Self-assembly)
284
Arrangement of Pd nanoparticles on SDS-functionalized single-walled carbon nanotubes
4. Conclusions
In summary, we report a route to site-selectively deposit
and arrange Pd nanopaticles on the surface of SDS-
functionalized SWCNTs. SDS formed supramolecular self-
assembly wrapped around SWCNTs. The terminal
functional group of SDS provided the combination site to
immobilize Pd nanoparticles. Pd nanoparticles are arranged
Fig. 2. TEM images of Pd nanoparticles selectively on the surface of SWCNTs and form ordered periodic
deposited on the SDS-functionalized SWCNTs. nanostructures. By control of the concentration of Pd (II)
ions, it also allows high density deposition of Pd
Figure 2 shows the high-resolution TEM images of nanoparticles on SWCNTs. This is very important for the
resulting nanostructures of Pd nanoparticles on the surface potential applications of these kinds of heterogeneous
285
Transactions of JWRI, Vol.39 (2010), No. 2
286
Transactions of JWRI, Vol.39 (2010), No. 2
†
Nano-carbon structures on silicon carbide
KEY WORDS: (Graphene) (Carbon nanotube) (Silicon carbide) (Transmission electron microscopy)
287
Nano-carbon structures on silicon carbide
were examined by observation using JEM-2010-type and is suggested that after the Si removal, on the Si-face,
EM-002B-type transmission electron microscopes. Thin carbon atoms make bonds parallel to the surface indicating
specimens for observation were made by an Ar-ion the formation of a two dimensional graphene structure,
thinning method. while on the C-face they make bonds perpendicular to the
surface, resulting in CNT formation.
3. Results and Discussion
Fundamentally, graphene layers were formed on Si- 4. Conclusions
terminated SiC (0001) surfaces [3-5], while CNTs were on Nano-carbon structures formed on the SiC surfaces
C-terminated (000-1) surfaces [6-8]. Figure 1 shows the were presented, and were investigated by HRTEM
HRTEM images of (a) graphene layers on SiC (0001) and observations. On the Si-terminated face of SiC,
(b) carbon-nanotube film on SiC (000-1). In Fig. 1(a), homogeneous graphene layers were obtained. Whereas
bi-layer graphene is observed as a dark line contrast. on the C-terminated face, highly-dense and well-aligned
Graphene layers are formed covering the whole terrace of CNTs were produced.
the SiC surface. In addition, similar line contrast can be
seen in the interface between graphene and SiC, which is References
denoted by a dashed line. Carbon atoms in this layer have [1] K.S. Novoselov, A.K. Geim, S.V. Morozov, D. Jiang, Y.
strong covalent bonds with silicon atoms beneath, and this Zhang, S.V. Dubonos, I.V. Grigorieva and A. A. Firsov: Science,
layer is called a buffer layer, which doesn’t have the 306 (2004), p.666.
[2] S. Iijima: Nature, 354 (1991), p.56.
electronic features of graphene. Elevating annealing [3] W. Norimatsu and M. Kusunoki: Chem. Phys. Lett., 468
temperature increases the number of graphene layers. (2009), p.52.
On the other hand, carbon nanotubes are formed on C- [4] W. Norimatsu and M. Kusunoki: Physica E, 42 (2010), 691.
terminated SiC as shown in Fig. 1(b). Characteristic [5] W. Norimatsu and M. Kusunoki: Phys. Rev. B, 81 (2010),
features of these carbon nanotubes are that almost all the p.161410.
well-aligned nanotubes are of zigzag-type and formed [6] M. Kusunoki, T. Suzuki, K. Kaneko and M. Ito: Phil. Mag.
perpendicularly to the SiC surface, and that the density of Lett., 79 (1999), p.153.
the nanotubes is extremely high. The thickness of the [7] M. Kusunoki, T. Suzuki, T. Hirayama, N. Shibata and K.
carbon nanotube film can be controlled by annealing Kaneko: Appl. Phys. Lett., 77 (2000), p.531.
[8] M. Kusunoki, T. Suzuki, T. Hirayama, N. Shibata and K.
temperature and time.
Kaneko: Chem. Phys. Lett., 366 (2002), p.458.
Common features of graphene and CNTs on SiC are [9] Z. Wang, S. Irle, G. Zheng, M. Kusunoki and K. Morokuma:
their homogeneity and their strong bonding with the J. Phys. Chem. C, 111 (2007), p.35.
substrate. These features and the selectivity of graphene
or CNTs are based on the crystal structure of SiC [9]. It
288
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Room temperature) (Coulomb oscillation) (Glassy alloys) (Nanoscale ) (Clusters) (Hydrogen)
(First-principles density functional calculation)
289
Room-temperature Coulomb oscillation of Ni-Nb-Zr-H glassy alloys with nanoscale size clusters
LQFUHDVHV DV WKH FXUUHQW GHFUHDVHV GRZQ WR ȝ$ DQG WKHQ Nb are nonmagnetic elements. In the atomic configuration
decreases in the lower current region from 100 to 1nA, of Ni 5 Zr 5 Nb 3 cluster, five Ni atoms are separately
indicating an occurrence of the optimum current for coordinated against Zr and Nb atoms.[8]
quantum dot tunnelling between the icosahedral Zr 5 Ni 5 Nb 3
cluster islands.
When we apply a magnetic field to the specimen from 0 References
to 15 T at 100 and 200 K in parallel and normal directions, [1] E.Ben-Jacob and Y.Grefen: Phys.Lett., 108A(1985), pp. 289.
we observed the voltage oscillation from around 6 T and 9 [2] M.A.Kastner, P.F.Kwasnick and J.C.Licini: Phys. Rev. B36
T at 100 K and 200 K, respectively. The amplitude of (1987), pp.8051.
vibration increases with increasing of magnetic field up to [3] For an example, H.W.Ch.Postma, T. Teepen , Z.Yao and M.
15 T except for a parallel magnetic field above ~12 T at Grifoni and C.Dekker: Science, 293 (2001), pp.76.
200 K. Here we should be noted that the chemical potential [4] M.Fukuhara, A.Kawashima, S.Yamaura and A.Inoue:
of the conduction electrons in the ferromagnetic island is Appl.Phys.Lett. 90 (2007), pp.20311.
affected by magnetic field. This is similar to variation in [5] M.Fukuhara, S.Yamaura and A.Inoue: J.Phys: Conf. Ser., 114
electrostatic potential caused by gate voltage in field (2009), pp.012086.
electrostatic transistor (FET). By analogy, we infer that [6] M.Fukuhara and A.Inoue: Europhy.Lett., 83 (2008), pp. 36002.
application of magnetic field induces Coulomb oscillation [7] M.Fukuhara and A.Inoue: J.Appl.Phys., 105 (2009), pp.
derived from noticeable variation in the chemical potential 063715.
of an isolated ferromagnet (island) in ferromagnetic fields. [8] M.Fukuhara, N.Fujima,H.Oji, A. Inoue and S.Emura: J.Alloy
In this study, the ferromagnetic metal responsible for the Comp., 497 (2010), pp.182.
magneto- Coulomb oscillation is Ni alone, because Zr and
290
Transactions of JWRI, Vol.39 (2010), No. 2
†
Superconductivity of Ni 0.324 Nb 0.216 Zr 0.36 H 0.1 glassy alloys
KEY WORDS: (Glassy alloys) (Superconductor) (Meissner effect) (Hydrogen charging) (Nanoclusters)
291
Superconductivity of Ni0.324Nb0.216Zr0.36H0.1 glassy alloys
References
[1] G.J. Fan and H. J. Fecht: J. Chem. Phys., 116 (2002), pp.5002.
[2] A.R. Yavari: Nature, 439 (2006), pp.405.
[3] M. Fukuhara, H. Yoshida, K. Koyama, A.Inoue and Y.miura:
J. Nanosci. Nanotechnol., vol.10 No.8 (2010).
[4] A.W. Smith, T.W. Clinton, C.C. Tsui and C. J. Lobb: Phys.
Rev. B, 49 (1994), pp.12927.
ᴾ [5] N. Kokubo, J. Aarts and P.H. Kes: Phys. Rev. B, 64 (2001),
pp.014507.
[6] W.L. Johnson, S.J. Poon, J. Durand and P. Duwez: Phys. Rev.
B, 18 (1978), pp.206.
[7]A. Inoue, Y. Takahashi, C. Suryanarayana and T. Masumoto: J.
Mater. Sci., 17 (1982), pp.1753.
[8] A. K. Bhatnagar, R. Pan and D.G. Naugle: Phys. Rev. B, 39
(1989) pp.12460
Fig. 3 (a) Magnetization vs. temperature [9] A. Inoue, C. Suryanarayana and T. Masumoto: J. Mater. Sci.,
curve under 50 Oe, (b) Magnetization loop 16 (1981), pp.1391.
at 2 K.
292
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Bulk metallic glass) (Dealloying) (Laser soldering) (Lead-free solder)
1. Introduction
Despite the good mechanical, chemical [1] and
magnetic [2] properties of bulk metallic glasses (BMGs),
the serious issue during the joint of BMG is the
recrystallization of the glassy phase at high temperatures. In
order to extend the engineering application of BMG it is
necessary to establish appropriate joint processes of
BMG/BMG and BMG/crystalline metal [3-4]. The lead-free
soldering process is a low temperature joining process and
therefore it can be performed below the crystallization
temperature of BMGs. However, many kinds of BMGs Fig. 1 Cross section of sample after HF treatment
such as Zr 55 Cu 30 Ni 5 Al 10 and Cu 60 Zr 30 Ti 10 exhibit no
wetting of lead-free solder on the surface of BMGs because In the reflow process, a Sn-3.0 mass%Ag-0.5 mass%Cu
of Zr- and Ti-oxide on their surface [5]. In previous (SAC) solder ball with a diameter of 1 mm was set on the
research, we found that selective leaching of Zr and Ti HF treated Cu 60 Zr 30 Ti 10 sample with RMA flux. Then, the
occurred upon immersing Cu 60 Zr 30 Ti 10 into HF solution, sample was put into a radiation furnace in a N 2 atmosphere
which resulted in the formation of a Cu rich layer [6]. and heated according to the temperature rise profile shown
The objective of this study is to evaluate wettability of in Fig. 2.
solder on the surface of HF treated Cu 60 Zr 30 Ti 10 in order to
make better joints between Cu 60 Zr 30 Ti 10 and lead-free
solder.
2. Experimental
Cu 60 Zr 30 Ti 10 BMG ribbon was cut into 20 mm ᶣ 20
mm ᶣ 70 ȝm strips and polished by SiC paper with 2000-
grit. The surface treatment was performed by immersing in
0.5 mol/L HF solution for 13 s at room temperature.ᴾ The
formation of Cu phase was confirmed with the amorphous
pattern by X-ray diffraction (XRD) as in the previous
study [6]. The porous layer with the thickness of about 500
nm was observed from scanning electron microscopy
(SEM) image as shown in Fig. 1. Fig. 2 Thermal profile of the reflow process.
Table 1 Some characteristics and preheating process of solder pastes used for laser soldering.
293
Enhancement of solderability of Cu60Zr30Ti10 bulk metallic glass by dealloying in hydrofluoric acid solution
294
Transactions of JWRI, Vol.39 (2010), No. 2
paste showed good wetting at broader laser power [2] A.Makino, T.Bitoh, A.Kojima, A.Inoue, T.Matsumoto.: J. of
ranges than SAC paste. The spread area of SB is larger ᴾ ᴾ Magnetism and Magnetic Mater., 215-216 (2000) , pp.288-292.
than that of SAC. This is attributed to the lower melting [3] Y.Kawamura: Mater. Sci. and Eng. A, 375-377 (2004),
temperature and slower Cu dissolution rate of SB. pp.112-119.
[4] Y.Kawamura, Y.Ohno: Scripta Mater., 45 (2001), pp.127-132.
In this study, the possibility to enhance the solderability of
[5] H.Nishikawa, K.WongPiromsarn, H.Abe, T.Takemoto,
Cu 60 Zr 30 Ti 10 by selective leaching in HF solution and laser ᴾ ᴾ M.Fukuhara, A.Inoue: Mater. Sci. and Eng. B, 148 (2008),
soldering was found. pp.124-127.
[6] H.Abe, K.Sato, H.Nishikawa, T.Takemoto, M.Fukuhara,
Reference A.Inoue: Mater. Trans., 50 (2009), pp.1255-1258.
[1]ᴾ A. Inoue: Materials Science and Engineering A, 267 (1999), ᴾ
ᴾ ᴾ pp.171-183.
295
Transactions of JWRI, Vol.39 (2010), No. 2
Local structures of amorphous (Ni 0.6 Nb 0.4 ) 100-x Zr x (x=30, 35, 40)
†
alloys by XAFS
MATSUURA Makoto *, KONNO Kazuya **, FUKUHARA Mikio *, FUJIMA Nobuhisa ***
and INOUE Akihisa ****
KEY WORDS: (XAFS) (Local structure) (Amorphous) (Metallic glass) (NiNbZr) (Icosahedral cluster)
1. Introduction the Zr and Nb K-edges and Si(111) for the Ni K-edge. The
A wide composition range of amorphous-forming ability higher harmonics of the incident X-ray were reduced by
has been reported in the ternary Ni-Nb-Zr alloy system by two Rh-coated mirrors. The XAFS data were analyzed
Kimura et al. [1]. Because Zr and Nb have high hydrogen using coded program Athena.
permeability, amorphous Ni-Nb-Zr alloys have been studied
to develop superior hydrogen permeation materials. Atomic 3. Results and discussion
structure of the amorphous Ni-Nb-Zr(-H) alloys has been Figure 1 shows the results of the XAFS oscillations ᷒
studied Yamaura et al [2]. and Sakurai et al [3] by XRD (k) weighted by k3 of the Ni-Kedges for the amorphous
and XAFS, respectively, in which the local structural (Ni 0.6 Nb 0.4 ) 100-x Zr x (x=30, 35 and 40) alloys. Figure 2
change by hydrogen absorption was the main issue. shows the results of the Fourier transform |F(r)| of k3Ὁ᷒(k)
Recently amorphous (Ni 0.6 Nb 0.4 ) 100-x Zr x (x=30 and 40) of the Ni K-edge for amorphous (Ni 0.6 Nb 0.4 ) 100-x Zr x (x=30,
alloys attracted much attention from a quiet different 35 and 40) alloys. The k-range in the Fourier transform is
viewpoint. Fukuhara et al. found strange behavior of the 2<k<12(Ḩ-1). The Fourier transforms of the Nb and Zr K-
electrical conductivity for the amorphous (Ni 0.6 Nb 0.4 ) 100- edges are shown in Fig. 3 and 4, respectively with the same
x Zr x H y (x= 30, 40 and y= 9 - 11) alloys such as the k-range as for the Ni K-edge.
Coulomb oscillation [4]. Oji et al. measured XAFS spectra The characteristic features of the XAFS results are that
for the amorphous (Ni 0.6 Nb 0.4 ) 100-x Zr x (x=30, 40) alloys the |F(r)| curves of the three elements do not change
and their hydrogenated (Ni 0.6 Nb 0.4 ) 100-x Zr x H y (x=30, 40 continuously with Zr concentration but change sharply from
and y=9 and 11) ones [5]. They achieved curve fitting x=30 to 35. This indicates that the local structures of the
analysis for the XAFS results of the Nb and Zr K-edges, amorphous (Ni 0.6 Nb 0.4 ) 100-x Zr x alloys of x<35 are different
(not for the Ni K-edge) and estimated lattice parameters for from those of x>35. The second feature of the XAFS results
these amorphous alloys and their hydrogenated one. is that the first peak of the |F(r)| for the Ni K-edge clearly
In order to elucidate the structural origin of the strange splits into two while those for the Nb and Zr ones show a
electrical behaviors for these hydrogenated amorphous single peak.
alloys, it is necessary to obtain more detail structure Considering atomic radii of the constituent elements, i.
information; concentration dependence of the local e. r Ni =1.24, r Nb =1.47 and r Zr =1.60 ( in Å ), the nearest
structures of the x=30, 35 and 40 for (Ni 0.6 Nb 0.4 ) 100-x Zr x neighbor peak of the |F(r)| for the Ni K-edge consists of Ni-
amorphous alloys and fitting analysis for the Ni K-edge
䢳䢲
XAFS data as well as the Nb and Zr K-edges. In the present (Ni0.6Nb0.4)100-xZrx
Ni K-edge
work XAFS measurements for the (Ni 0.6 Nb 0.4 ) 100-x Zr x 䢺
x=40
(x=30, 35 and 40) alloys were carried out and the results
were compared with the first principle calculation based on 䢸
䢶
x=35
䢵
2. Experiments 䢴
x=30
Ribbon samples were prepared by a melt-quenching
䢲
technique using a single-roller apparatus. XAFS
measurements of the Ni, Nb and Zr K-edges were made at 䢯䢴
Hyogo, Japan. The incident X-ray was monochromatized Fig.1 ;$)6 RVFLOODWLRQ ȤN 㺃 k3 of the Ni K-edge for
by a Si double crystal monochromator: Si(311) plane for amorphous (Ni0.6Nb0.4)100-xZrx (x=30, 35 and 40) alloys at 20K.
† Received on 30 September 2010 ****ᴾ Tohoku University,2-1-1 Katahira, Aoba,Sendai 980-8577
* Institute for Materials Research, Tohoku University, Japan
2-1-1 Katahira, Aoba, Sendai 980-8577, Japan Transactions of JWRI is published by Joining and Welding
** National College of Technology-Sendai, 48 Nodayama, Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
Shiote-Aza, Natori 981-1239, Japan Japan
*** Faculty of Engineering,Shizuoka University,Hamamatsu
432-8561, Japan
296
Local structures of amorphous (Ni0.6Nb0.4)100-xZrx(x=30, 35, 40) alloys by XAFS
䢺 䢵䢰䢷
䢴䢰䢷
䢢䣺䢿䢵䢲
䢢䣺䢿䢵䢷
䣾䣈䢪䣴䢫䣾䢢䣱䣨䢢䣭 䢬䢢ȮN
䢸
䢴䢰䢲
䢢䣺䢿䢶䢲
䢵
䢳䢰䢷
x=40
䣾䣈䢪䣴䢫䣾䢢䣨䣱䣴䢢䣭 䢬ȮN
䢳䢰䢲
䢵
䢶
䢲䢰䢷
䢲䢰䢲
x=35 䢲䢰䢷 䢳䢰䢲 䢳䢰䢷 䢴䢰䢲 䢴䢰䢷 䢵䢰䢲 䢵䢰䢷 䢶䢰䢲
䣴䢪ύ
䢴
Fig.3 Fourier transform of the Nb K-edge of k3㺃
ȤN IRU the amorphous (Ni0.6Nb0.4)100-xZrx
x=30 (x=30, 35 and 40) alloys at 20K.
䢲
䢲䢰䢷 䢳䢰䢲 䢳䢰䢷 䢴䢰䢲 䢴䢰䢷 䢵䢰䢲 䢵䢰䢷 䢶䢰䢲
䣴䢪ύ 䢷
Fig.2 Fourier transform of the Ni K-edge of k3㺃ȤN (Ni0.6Nb0.4)100-xZrx
Zr K-edge
for the amorphous (Ni0.6Nb0.4)100-xZrx (x=30, 35 and 䢶
䣾䣈䢪䣴䢫䣾䢢䣱䣨䢢䣭 䢬ȮN
Ni correlations and the second one Ni-Nb and Ni-Zr ones. 䢵 x = 40
䢵
This can be proved by the fact that the intensity of the first
shoulder in the |F(r)| decreases with increase in Zr 䢴
297
Transactions of JWRI, Vol.39 (2010), No. 2
CHO Ken *,****, TAKENAKA Kosuke *,****, SETSUHARA Yuichi *,****, SHIRATANI
Masaharu **,****, SEKINE Makoto ***,**** and HORI Masaru ***,****
KEY WORDS: (Inductively-coupled plasma), (Low-inductance antenna), (Low-damage process), (Hard x-ray
photoelectron spectroscopy)
1. Introduction
Next-generation devices including flexible displays, generator at 13.56 MHz via a matching network. Each LIA
thin-film photovoltaic cells and advanced ULSIs (low-k unit was connected in parallel to the matching network. The
interlayer and EUV lithography) are expected to occus in LIA unit consists of a U-shaped antenna conductor, which
combination with polymer materials [1,2]. For successful was fully covered with dielectric tubing for complete
development of these devices, therefore, it is of great isolation from the plasma [4]. The chamber had a 500 mm
significance to control polymer surface and/or interface inner diameter and a 200 mm height, which was connected
with functional layers in terms of chemical and physical to a diffusion chamber with a 500 mm inner diameter and a
structures. In order to satisfy these requirements, it is 400 mm height. The Ar-O 2 mixture plasma was sustained at
necessary to use plasma processes, because plasma has an oxygen partial pressure (R p ) of 20% at a total pressure of
high-reactivity due to ions and radicals. However, bond- 2.6 Pa at an RF power of 1 kW. Optical emission spectra in
dissociation energies of polymer materials for thermal UV and VUV regions were measured with a VUV
decomposition are less than that of inorganic materials (3.4 monochrometer mounted on the sidewall of the chamber.
eV for the O-C(=O) bond, 3.6 eV for the C-CH 3 bond and For evaluating the interactions of the polymer surface
8.4 eV for C=O bond). Thus, it is considered that ions, with UV and VUV photons from the plasmas, PET films
radicals, photons and electrons from plasma may (thickness 100 Pm) were exposed to photons from the Ar-
give significant degradation in chemical bonding structures O 2 mixture plasmas by covering the PET samples with a
of soft materials beneath the surface and/or the MgF 2 window. Here it is noted that thickness of the MgF 2
inorganic/organic interface. Therefore, it is necessary to window is 2 mm. The MgF 2 window is transparent in
understand plasma-polymer interactions for development of UV+VUV regions (wavelength longer than 115 nm).
low-damage and low-temperature plasma processes. Therefore, it is possible to evaluate the effects of photo-
As one of the effective plasma sources for these irradiation by comparing results of the sample with a
processes, inductivity-coupled plasma sources have been plasma exposed sample [5].
developed with multiple low-inductance antenna (LIA)
modules, which allowed low-voltage operation of high-
density ICPs [3]. In terms of interactions of radiations from
plasmas with organic materials surface, radiations in
vacuum ultraviolet (VUV) and ultraviolet (UV) are
significant in that the photon energies are higher than the
bond dissociation energies. In this study, interactions of
optical emissions from Ar-O 2 mixture plasmas with
polymer surfaces have been investigated on the basis of
surface morphological changes and non-destructive depth
analyses of chemical bonding states in the nano-surface
layer of polyethyleneterephthalate (PET) films via hard x-
ray photoelectron spectroscopy (HXPES).
2. Experimental
A schematic diagram of the process chamber is shown
in Fig. 1. The LIA units were mounted on the top flange of Fig. 1 Schematic diagram of the discharge chamber.
the discharge chamber and were coupled to a RF power
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* JWRI, Osaka Univ., Osaka, Japan Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
** Dpt. of Electronics, Kyushu Univ., Fukuoka, Japan Japan
*** PLANT, Nagoya Univ., Nagoya, Japan
**** JST, CREST
298
Effects of photon irradiation in UV and VUV regions during plasma processing of organic materials
Fig. 2 Typical AFM images of the PET surface; (a) before and (b) after direct exposure to the plasma,
(c) after irradiation with the UV+VUV photons from the plasma.
For investigation of surface morphological changes, the The HXPES and XPS C1s spectra for PET films
sample surface was observed with an atomic force without and with direct exposure to the Ar-O 2 mixture
microscope (AFM). The depth analysis of chemical plasma and irradiations with photons from the Ar-O 2
bonding states was carried out by hard x-ray photoelectron mixture plasma are shown in Fig. 3. The HXPES spectra
spectroscopy (HXPES) and conventional x-ray showed insignificant change of the O=C–O bond and the
photoelectron spectroscopy (XPS). The HXPES analysis C–O bond after the plasma exposure and the UV+VUV
was performed with x-rays from a synchrotron at a photon photo-irradiation. Whereas, in the XPS spectra, the samples
energy of 7940 eV at the national synchrotron radiation exposed to the Ar-O 2 mixture plasma and the UV+VUV
facility (SPring-8) of the Japan Synchrotron Radiation photons clearly exhibit significant increase in the oxygen
Institute [6]. The inelastic mean free path (IMFP) of 7940 functionalities (the O=C–O bond and the C–O bond). In
eV electrons in polymers is approximately 18 nm [7]. particular, substantial increase is found for the C=O bond,
Therefore, the probing depth of HXPES is about 50 nm for which is not observed in the spectrum of raw PET film.
analysis of polymers The XPS analysis was performed with The UV+VUV photo-irradiation from the Ar-O 2
non-monochromatized Mg KD radiation (photon energy of mixture plasmas resulted in insignificant change of the bond
1253.6 eV). structures in deeper regions up to about 50 nm from the
surface. However, in regions up to about 10 nm from the
3. Results surface, the UV+VUV photo-irradiation resulted in increase
Figure 2 shows typical AFM images of the PET surface of the oxygen functionalities.
(a) before and (b) after direct exposure to the plasma, (c)
after irradiation with the UV+VUV photons from the 4. Summary
plasma (with MgF 2 window). Surface roughness R q (root- Interactions of UV+ VUV photons from Ar-O 2 mixture
mean-square) of the raw PET film (Fig. 2(a)) was 1.0 nm. plasmas with polymer surfaces were investigated using
After the direct exposure to the Ar+O 2 mixture plasma for AFM, HXPES and conventional XPS analyses. The AFM
1052 seconds, the roughness increased to 5.6 nm (Fig. 2 images indicated that the surface morphological change of
(b)). Whereas, the roughness was almost constant (R q = 0.9 PET films due to Ar-O 2 mixture plasma exposure is not
nm) even after exposure to UV+VUV photons from the caused by only UV+VUV photons from the plasma. The
plasma for 1052 seconds as shown in Fig. 2(c). These HXPES analysis suggested that the effects of UV+VUV
results suggest that the surface morphological change of photons from the Ar-O 2 mixture plasmas on the chemical
PET films due to the Ar-O 2 mixture plasma exposure is not bonding states of polymer are insignificant in deeper
caused by only UV+VUV photons from the plasma. regions up to about 50 nm from the surface. The XPS
analysis shows that the formation of oxygen functionalities
HXPES, TOA = 80deg
after UV+VUV photo-irradiation is localized in the vicinity
XPS
of the surface regions as shallow as 10 nm. These results
O=C-O (288.6eV)
C=O (287.4eV)
C-O (286.1eV)
O=C-O (288.6eV)
C=O (287.4eV)
C-O (286.1eV)
C-C (284.6eV)
C-C (284.6eV)
˭ය˭
(291.2eV)
˭ය˭ (291.2eV)
C1s C1s
Fig. 3 HXPES and XPS C1s spectra for PET films; Acknowledgement
(a) before and (b) after direct exposure to the plasma, This work was partly supported by The Global COE
(c) after irradiation with the UV+VUV photons from the Program "Center of Excellence for Advanced Structural
plasma. and Functional Materials Design" through the Ministry of
299
Transactions of JWRI, Vol.39 (2010), No. 2
Education, Culture Sports, Science and Technology [3] Y. Setsuhara, T. Shoji, A. Ebe, S, Baba, N. Yamamoto, K.
(MEXT), Japan, and by a Grant-in-Aid for Young Takahashi, K. Ono and S. Miyake, Surf. Coat. Technol. 174 -
Scientists (B) from the Japan Society for the Promotion of 175 (2003) 33.
Science (Research Project Number: 21760587). [4] Y. Setsuhara, K. Takenaka, A. Ebe and K. Nishisaka, Plasma
Process. Polym. 4 (2007) S628.
References [5] S. Uchida, S. Takashima, M. Hori, M. Fukasawa, K.
[1] M.-C. Choi, Y. Kim and C.-S. Ha, Prog. Polym. Sci. 33 Ohshima, K. Nagahata, and T. Tatsumi, J. Appl. Phys. 103,
(2008) 581. 073303 (2008)
[2] K. L. Chopra, P.D. Paulson and V. Dutta, Prog. Photovolt: [6] K. Kobayashi, Nucl. Instrum. Meth. Phys. Res. A 601 (2009)
Res. Appl. 12 (2004) 69. 32.
[7] S. Tanuma, C.J. Powell and D.R. Penn, Surf. Interface. Anal.
21 (1994) 165.
300
Transactions of JWRI, Vol.39 (2010), No. 2
†
Single layer PDMS flexible parallel wall microvalvets
2. Principle
Figure 1 shows a working principle of single droplet
sampling by horizontal valves and flexible parallel walls.
PMDS walls on both sides of the micro channel are
Fig. 2 The principle of multi modes droplet sampling.
deformed by controlled air pressure. The deformation,
especially, overshot deformation of valves makes a
sequential deflection of the flexible walls inside the channel.
Each fluidic channel is designed to have the same flow
301
Single layer PDMS flexible parallel wall microvalvets
main stream out of the device under initial condition. remaining issues. It is expected that integration of an
Designed structures were formed from PDMS using a SU-8 automatic pressure control system will be a solution of
mold. It is bonded to the PDMS coated glass substrate after these problems. Some specific structures which can realize
plasma pretreatment. The fabrication results by single step independent control of droplet generation from sampling
SU-8 patterning and PDMS replication are shown in Fig. 4. are considered.
In order to realize flexible PDMS structures, resin and
curing agent were mixed in a 15:1 ratio. For fluidic
experiments, syringe (1750CX, HAMILTON) and syringe
pump (KDS210, kdScientific) were used. The air pressure
was controlled by pressure regulator (2657 pneumatic
pressure standard, YOKOGAWA).
Also, a CCD camera (JK-TU53H, TOSHIBA) and a data
processing computer were utilized for visualization and
storage of the droplet sampling processes.
(a) 0 sec (b) 1 sec
302
Transactions of JWRI, Vol.39 (2010), No. 2
303
Transactions of JWRI, Vol.39 (2010), No. 2
1. Introduction of the IR beam was set to be 60° from the surface normal.
Iontophoresis is a drug delivery method that can be used
to enhance the movement of ionic compounds or non- 3. Results and Discussion
charged drug molecules by the application of external Figure 1 shows the FT-IR spectra of the Al electrodes
electric fields. Iontophoresis can transport ionic substance subjected to the electrochemical deposition process of
across skin or membranes into the body. Although alginic acid under the current density of 1 mA/cm2 for 120 s,
iontophoresis needs a drug-loaded electrode, a metal 300 s, and 600 s. The characteristic peaks at 1420 cm-1 and
electrode is unable to adsorb a large amount of drugs. Thus, 1600 cm-1 were assigned to dissociated carboxyl groups (–
the electrode that conjugates drug-absorbable gels is COO-) of alginate, which corresponded to the symmetric
desired for sustainable administration. Alginic acid and and asymmetric C=O stretching vibration. The result
chitosan have been widely studied as drug delivery carriers. indicated that the electrochemical process successfully
Sodium alginate or chitosan form gels around electrodes by deposited the alginate on the electrodes. The deposited
the pH change of the solution that was generated by mass on the electrodes was increased with an increase of
electrolysis of water. In this study, we investigated the the applied times. The band at 1740 cm-1 was assigned to
interface stability of the gels formed on a metal electrode by C=O stretching vibration of the undissociated carboxyl
applying direct currents in the solutions in order to develop groups (–COOH). It was already described that alginic acid
a novel iontophoresis electrode. gels were formed around anode electrodes due to
electrochemical decomposition of water [1]. According to
2. Materials and Methods the literature, the related reactions at the anode were
Al electrodes were cut into 2.5 x 1 x 0.01 cm pieces. expressed as follows;
The electrodes were ultrasonically washed three times with
acetone for 10 min, and then dried under nitrogen gas flow. H 2 2ĺ++ + 1/2O 2 + 2e- (1)
The cleaning procedure was completed by 10 min
UV/ozone irradiation (ProCleaner 110, BioForce Application of an electric field caused a pH decrease at the
Nanosciences, Inc., USA). The electrochemical deposition anode. In the aqueous solution, sodium alginate was
cell was a two-electrode system, which consisted of a dissociated to sodium ions and alginate molecules, and then
cylindrical polystyrene cell and ᴾ metal electrodes. The the alginate molecules were protonated around the anode.
deposition was obtained on Al electrodes that were set as
anodes for alginic acid and as cathodes for chitosan. The Pt Na-$OJĺ$OJ- + Na+ (2)
electrode was used as a counter electrode. The distance Alg- + H+ ĺ+-Alg (3)
between two electrodes was 10 mm. Twenty mL of 0.5%
sodium alginate aqueous solution or 0.2% chitosan solution where the symbols Na-Alg and H-Alg denote sodium
was poured into the cylindrical polystyrene cell. The alginate and alginic acid, which are strong and weak
surface area of the electrodes exposed to the solution was electrolytes, respectively. The formed H-Alg turned to gels
adjusted to be 4 cm2. The electrochemical deposition was via hydrogen bonding between –COOH of H-Alg, which
conducted under application of a DC electric field with was confirmed in the FT-IR spectra shown in Fig. 1. Thus,
current densities of 1.0-5.0 mA/cm2 for up to 3600 s at the alginic acid layers were deposited through the proton-
room temperature. The electrodes were dried at 37°C for 24 mediated gelation. The intensity of the peaks for –COOH at
h. The surface of the electrodes was characterized by FT-IR. 1740 cm-1 was increased with an increase of the applied
All measurements were performed using FT-IR times. This result indicated that the prolonged application
spectrometer, model FT/IR-4100 (JASCO, Japan) equipped leads to an increase of the hydrogen bonding in the layers in
with PR-5101 system (JASCO, Japan). The incident angle addition to an increase of the deposited amount.
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* Tokyo Institute of Technology, Tokyo, Japan Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
** National Institute for Materials Science, Tsukuba, Japan Japan
*** Tokyo Medical and Dental University, Tokyo, Japan
**** TTI ellebeau, Inc., Tokyo, Japan
304
Development of a novel integrated iontophoresis electrode consisting of metallic and drug-loaded layers
4. Conclusion
Drug-absorbable gels of alginic acid or chitosan were
successfully deposited on the Al electrodes.
References
[1] F. Yokoyama et al.: Eur. Polym. J., 34 (1998), pp.229-234.
[2] G. A. Parks: Chem. Rev., 65 (1965), pp. 177-198.
305
Transactions of JWRI, Vol.39 (2010), No. 2
†
Responses of osteoblast to surface modified bulk metallic glass
KEY WORDS: (Metallic glass) (Femtosecond laser) (Hydrothermal-electrochemical treatment) (Osteoblast) (Cell
attachment assay) (Yinculin expression)
1. Introduction kHz, respectively. The laser beam was focused onto the Ti-
Metallic glass has many attractive properties as a based BMG (Ti 40 Zr 10 Cu 36 Pd 14 ) surface by a lens with a
biomaterial, such as high strength, high elastic limit, low 100 mm focal length. The Gaussian laser beam had a
Young’s modulus and high corrosion resistance. Surface GLDPHWHURIȝPDWWKH/e2 intensity points) on the Ti-
modifications of the metallic materials have been tried due based BMG surface. Ti-based BMG’s surface was
to insufficiency of biocompatibility. One of the techniques irradiated with the femtosecond laser at the average laser
is a combination of femtosecond laser and hydrothermal- fluence of 0.15 and 1.0 J/cm2, respectively. The laser beam
electrochemical treatment (Ref. 1,2). The modified surfaces was scanned on the Ti-based BMG’s surface in the area of
are expected to obtain bioactive property since apatite 5 x 5 mm2 using the XY stage. Then, laser scanning speed
forming ability in the simulated body fluid has been was 1 mm/sec and hatching distance was 20 μm.
demonstrated. To evaluate biocompatibility in vitro, the
objective study was to examine cellular behavior c) Hydrothermal-electrochemical (H-E) treatment
responding to the surface modified metallic glass. The After femtosecond laser irradiated, Ti-based BMG was
attachment, spread, and adhesion of osteoblasts on the subjected to H-E treatment at 90 °C for 120 hours in 5
materials are investigated. mol/L NaOH solutions to attain the surface bioactivity. The
cylindrical Ti-based BMG (Ti 40 Zr 10 Cu 36 Pd 14 ) specimen, 5
2. Materials and Methods mm in diameter was used as the working electrode facing
1. Fabrication of surface modified Ti-based bulk metallic the circular surface to the counter electrode of platinum
glass sheet, 10×50×0.5 mm3. The constant dc voltage of 1.4V
a) Ti-based (Ti 40 Zr 10 Cu 36 Pd 14 ) bulk metallic glass (BMG) was applied between these electrodes keeping their
samples distance at 4 cm during H-E treatment.
The Ti 40 Zr 10 Cu 36 Pd 14 alloy ingots (compositions are
given in nominal atomic percentages) were prepared by arc 2. Cell culture and evaluations of biological responses
melting mixtures of pure Ti, Zr, Cu, and Pd metals (>99.9 a) Cell culture
mass%) in a high purity argon atmosphere purified using Zr Osteoblast-like MG63 cells were seeded directly onto
getter. Bulk cylindrical rods with a diameter of 5 mm were the Ti-based BMG’s surface irradiated with the
prepared by melting of the master ingots followed by tilt femtosecond laser at the average laser fluence of 0.15 and
casting into copper molds. 1.0 J/cm2 and subjected H-E treatment (designated 0.15
The disc samples with 5 mm in diameter and 2 mm in and 1.0 J/cm2 surface, respectively).
height were cut out from the rods, then polished. The
glassy structure of as-prepared samples was examined by b) Cell attachment assay and morphology analysis
X-ray diffractometry in reflection with monochromatic Cu The attachment of cells were fixed and stained with
KĮ radiation. The thermal stability was examined by fluorescent dye, rhodamine phalloidin and DAPI, and
differential scanning calorimetry (DSC) under an argon analyzed using a microscopic image-based observation.
atmosphere at a heating rate of 0.67 K/s.
c) Vinculin expression analysis
b) Femtosecond (Ti:sapphire) laser The expression and localization of focal adhesion
A femtosecond (Ti:sapphire) laser was employed in this protein, vinculin, were analyzed using a microscopic
experiment. The wavelength, pulse duration and repetition image-based observation and densitometry.
rate of the femtosecond laser were 775 nm, 150 fs and 1
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* IBB, Tokyo Medical and Dental University, Tokyo, Japan Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
** MSL, Tokyo Institute of Technology, Tokyo, Japan Japan
*** JWRI, Osaka University, Osaka, Japan
**** IMR, Tohoku University, Miyagi, Japan
***** Tohoku University, Miyagi, Japan
306
Responses of osteoblast to surface modified bulk metallic glass
References
Fig. 2 Representative microscopic images of cells stained [1] M. Fukuhara, X. Wang and A. Inoue: Appl. Phys. lett., 91
with rhodamine phalloidin for actin filaments (red) and (2007), p.171908.
DAPI (blue on (A) 0.15 and (B) 1.0 J/cm2 surface. [2] M. Yoshimura, T. Onoki, M. Fukuhara, X. Wang, K. Nakata
and T. Kuroda: Mater. Sci. Engineer. B, 148 (2008), pp.2-6.
307
Transactions of JWRI, Vol.39 (2010), No. 2
OHARA Satoshi *, SATO Kazuyoshi **, TAN Zhenquan *, UMETSU Mitsuo ***
308
Joining of ceramic nanocrystals and bio-molecules towards bio-medical applications
(a)
(b) 10 nm
Fig. 2 TEM image of a zirconia nanoparticle assembly by
O-DNA.
309
Transactions of JWRI, Vol.39 (2010), No. 2
†
Cu-based metallic glass surfacemodified with Cu for soldering
Productions and applications of Cu clad metallic glass
KEY WORDS: (Cu-based metallic glass) (Clad material) (Soldering) (Wettability) (Quench) (passive oxide
film) (Surface modification)
Fig. 1 Production of Cu thin film cladding bulk Fig. 2 Cross-section image of Cu-cladding bulk
metallic glass (BMG). metallic glass Cu36Zr48Al8Ag8.
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* JWRI, Osaka University, Ibaraki, Japan Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
** IMR, Tohoku University, Sendai, Japan Japan
*** Tohoku University, Sendai, Japan
310
Cu-based metallic glass surfacemodified with Cu for soldering Productions and applications of Cu clad metallic glass
oxide film on the metallic glass. However, the removable phase identification was conducted by micro-focused X-ray
area was limited around the ultrasonic-transducer. diffraction (XRD) analysis using Co KD radiation. The
In this paper, we report a development of Cu thin film microstructure of the interface between Cu thin film and
clad Cu-based metallic glass Cu 36 Zr 48 Al 8 Ag 8 for the aim of metallic glass was observed with an optical microscope
improving solder wetting. By cladding Cu thin film over the (OM), a scanning electron microscope (SEM), and a
transmission electron microscope (TEM). The Surface
oxide layer was analyzed by auger electron spectroscopy
(AES). Solder wettability was evaluated by dipping Sn-
3Ag-0.5Cu solder on the Cu thin film clad metallic glass at
553K for 60 s.
311
Transactions of JWRI, Vol.39 (2010), No. 2
ATOU Toshiyuki *, MUKOGAWA Shintaro **, KIMURA Hisamichi ***, ITO Shun ***
and KIKUCHI Masae ****
KEY WORDS: (Metallic glass) (Ceramics) (Joining) (Shock wave) (High pressure) (High temperature)
(Short duration)
2. Experimental
Shock-loading experiments were performed using a
two-stage light gas gun at Tokyo Institute of Technology up
to 2.1 km/s (Fig. 1(a)). 3-mm-thick and 21-mm –diameter
Cu flyer plates were used. Flyer velocity was measured
using a magneto-flyer method, and the pressures in the
recovery capsule were determined by an impedance match
method [3]. Fig. 1(b) shows a shock recovery system used
for recovery experiment. Zr 55 Al 10 Ni 5 Cu 30 metallic glass
was formed into disk 14-mm in diameter and 1-mm in
thickness and then encased in an iron recovery capsule with
quartz (SiO 2 ) powder in 1-mm thick. Porosity of quartz
powder was estimated to be about 50 %. The recovery
capsule was placed within a steel recovery fixture [4].
Recovered specimens were examined using SEM, TEM, Fig. 1 (a) Two-stage light gas gun installed at Tokyo
and EPMA measurements to investigate the microstructures Tech (b) Shock recovery system used for shock
in the interface between metallic glass and SiO 2 . assisted joining experiments (see text).
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* MSL, Tokyo Tech., Yokohama, Japan Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
** IGSSE, Tokyo Tech.,, Yokohama, Japan Japan
*** IMR, Tohoku Univ., Sendai, Japan
**** Kansei Center, Tohoku Fukushi Univ., Sendai, Japan
312
Shock-assisted joining between metallic glass and ceramics
6L2 =UEDVHGEXONPHHWDOOLF
JODVV
wP
Fig. 2 SEM photograph of interface of metallic glass Fig. 4 TEM photograph of interface of metallic glass and
and SiO2 shock-loaded to 24 GPa. amorphous SiO2 shock-loaded to 28 GPa.
4. Conclusions
The conclusions of this study are summarized as follows.
(1) Joining between metallic glass and ceramic was tried
using impulsive shock-loading to Zr 55 Al 10 Ni 5 Cu 30
1L 6L metallic glass and quartz (SiO 2 ) system.
(2) We have confirmed that metallic glass could be covered
with thin layer of SiO 2 .
(3) Microtexture observation suggested that adhesion
mechanism might be different against applied shock
pressure.
References
[1] M. Kikuchi, Y. Syono, K. Fukuoka, K. Hiraga: J. Mat. Sci.
Lett., 6 (1987), pp.97-99.
wP
[2] R. Murao, M. Kikuchi, K. Fukuoka, E. Aoyagi, T. Atou, Y.
Syono: J. Mat. Res., 14 (1999), pp.3169-3174.
Fig. 3 Mapping by EPMA of interface of glass and SiO2 [3] T. Goto, Y. Syono: Materials science of the Earth’s Interior
shock-loaded to 24 GPa. (Ed: I. Sunagawa), Terra, Tokyo (1984), pp.605.
[4] T. Hongo, T. Atou, S. Ito, K. Yubuta, M. Kikuchi, K. G.
These results suggest that the joining of this interface might Nakamura, S. Itoh, K. Kusaba, K. Fukuoka, and K. Kondo:
be due to physical (mechanical) adhesion, as distinct from Phys. Rev, B76 (2007), pp.104114.
the chemical reaction between metallic glass and quartz, as
observed in the previous shock experiments [1, 2]. On the
other hand, recovered sample from 48 GPa showed metallic
313
Transactions of JWRI, Vol.39 (2010), No. 2
314
Development of plasma MIG brazing process of advanced materials
References
[1] T. Kozuru et al: Preprints of the national meeting of J.W.S.,
83 (2008) pp.330-331.
315
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Steel) (Phase transformation) (Time-Resolved X-ray Diffraction) (Martensite) (LSCM)
316
Developments of hybrid in-situ observation system to study the microstructural change of metallic alloys
317
Transactions of JWRI, Vol.39 (2010), No. 2
4. Conclusions [2] S.S. Babu, J.W. Elmer, J.M. Vitek, S.A. David: Acta Mater,
The detailed configuration of the hybrid TRXRD/LSCM 50 (2002), pp. 4763-4781.
system was explained. The phase transformation behavior [3] H. Yin, T. Emi, H. Shibata: Acta Mater, 47 (1999), pp. 1523-
of low-temperature transformation steel was traced by using 1535.
[4]Y. Komizo and H. Terasaki: Sci. and Tech. Weld.
the system and it was shown that the hybrid system enable
Joining,(2010)doi:10.1179/136217110X12785889549822.
in-situ observation of phase trasnformation of metallic [5] S.S. Babu, E.D. Specht, S.A. David, E. Karapetrova, P.
alloys in real and reciprocal lattice space. Zschack, M. Peet, H.K.D.H. Bhadeshia: Metall Mater Trans A,
36A (2005), pp. 3281-3289.
References [6] H. Terasaki and Y. Komizo: Scripta Mater., 64 (2011) pp. 29-
[1] J. M. Vitek, A. Dasgupta and S. A. David: Metall. Mater. 32.
Trans. A, 14 (1983), pp. 1833-1841.
318
Transactions of JWRI, Vol.39 (2010), No. 2
SAKINO Yoshihiro *, IZUMI Yui **, KURODA Toshio *, SAKAGAMI Takahide ***
and KIM You-Chul *
KEY WORDS: (Zr-based metallic glass) (High speed tensile test) (Thermography) (Fracture) (Spark)
ᴾ ᴾ ᴾ ᴾ ᴾ ᴾ ᴾ ᴾ ᴾ ᴾ (Hot oxide particle) (Liquid droplet)
319
Measurement of fracture behavior of Zr-based metallic glass by thermograpy
based metallic glass. and the fracture surface was heated to high temperatures,
The mechanism generating this sparking phenomenon which caused oxidation.
has been suggested as follows: (1) thermal insulation From these observations, the flying particles, that are
phenomenon, (2) spattering of hot oxide particles, (3) observed as the sparks, are hot oxide particles and liquid
spattering of liquid droplets, (4) process of vanishing or droplets.
elimination of excess electrons specific to metallic glass,
and so on. This sparking phenomenon seems to be unique 4. Conclusions
to amorphous metallic glass. The conclusions of this study are summarized as follows.
Considering that the flying particles are also attached to (1) The sparking phenomenon was also observed in
the fracture surface, we examined the fracture surface using Zr 48 Cu 36 Al 8 Ag 8 and Zr 55 Cu 30 Al 10 Ni 5 metallic glass
SEM observation and XRD analysis. Figure 4 (a) and (b) specimens regardless of strain rate at the moment of
shows the fracture surface of Zr 48 Cu 36 Al 8 Ag 8 metallic fracture. However the amount of sparks are different,
glass in the tensile test at a low and high strain rate.
Regardless of the strain rate, smooth dimples (A) and
smooth areas (B) are observed, in addition to the rough
dimple fractures normally observed that indicate ductile
fracture. The smooth dimples (A) are regarded as attached
marks of flying particles due to sparking. These indicate
that the hot oxide particles collided with the fracture surface
and become attached. The smooth areas (B) are regarded as
melting and solidification cracking. This indicates that it
was heated to the melting temperature of metallic glass,
which is 900°C or higher, by thermal insulation change, etc.,
during the tensile fracture of the metallic glass, causing it to
melt, and then generate high-temperature cracks during
cooling. The large localized deformation and the remelting
feature are believed to be due to the high strength, large
elastic energy and low melting temperature of the alloy,
which are common characteristics of bulk metallic glass.
As for the results of XRD analysis, most of the fractures
had a halo pattern indicating an amorphous structure.
However, a peak for ZrO 2 oxide film or oxide particles or a
crystallization peak for substances such as ZrNi 2 were
observed in some parts along with the halo patterns. Thus,
some of the fracture particles become attached to the
fracture surface as flying particles at the moment of fracture Fig. 4 Fracture surfaces of tensile test. (a) Crosshead
speed 0.001 mm/s. (b) Crosshead speed 1200 mm/s.
320
Transactions of JWRI, Vol.39 (2010), No. 2
depending on the components of Zr-based metallic glass. with the experiment. This work was supported by Grant-in-
Aid for Cooperative Research Project of Nationwide Joint-
(2) It is revealed by SEM observation and XRD analysis of
Use Research Institute on Development Base of Joining
the fracture surface that the flying particles of hot oxide
Technology for New Metallic Glasses and Inorganic
particles and liquid droplets are observed as sparks at
Materials from The Ministry of Education, Culture, Sports,
the moment of fracture. Although this can only be seen
Science and Technology, Japan.
as a momentary spark when observed visually, it could
be detected with an IR thermograph camera. References
[1] G.Y. Wang, P.K. Liaw, W.H. Peter, B. Yang, Y. Yokoyama,
Acknowledgements M.L. Benson, B.A. Green, M.J. Kirkham, S.A. White, T.A.
The authors sincerely thank Professor Inoue and Saleh, R.L. McDaniels, R.V. Steward, R.A. Buchanan, C.T.
Associate Professor Kimura, Tohoku University for Liu and C.R. Brooks: Intermetallics, 12 (2004), pp.885-892.
supplying the samples for this study.ᴾ The authors also [2] Y. Sakino, T. Kuroda, A Kobayashi and Y.-C. Kim: Frontier
thank Technical Specialist Mr. Nakatsuji, for his assistance of Applied Plasma Techmology, 2(3), pp.102-106.
321
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Plasma chemical vapor deposition) (Amorphous boron carbon oxynitride) (Diaphragm)
(Environmental cell) (Transmission electron microscope)
1. Introduction chamber to isolate the vacuum around the TEM from the
It is an unquestionable fact that bulk gold is chemically reaction gas atmosphere around the specimen by means of a
inert. Recently however, it was reported that gold particles diaphragm [3]. The diaphragm is the most important part of
of less than 10 nm in diameter exhibit unique catalytic the E-cell TEM system because it not only maintains the
properties [1]. The level of catalytic activity is particularly pressure difference between the vacuum and the reaction
high in the oxidation of carbon monoxide (CO) and similar gas but also transmits the electron beam. Amorphous
substances when the particles are bound tightly to specific carbon thin films have so far been used as the diaphragm to
metal oxides, such as TiO 2 [2]. To determine the reaction allow observation of the oxidation of CO using the gold
mechanism of this gold nanoparticle catalyst, which is nanoparticle catalyst. The carbon thin film (thickness 8 nm)
unclear, requires investigation of the gold nanoparticle has advantages such as high hardness, no electron
surface and interface atomic structures. Transmission diffraction contrast and low electron scattering. However,
electron microscopy (TEM) is useful for analyzing extended observation is not possible as the film suffers
structures at the atomic level, but as specimens are normally damage from the electron beam and the oxygen in the E-
placed under a strong vacuum, it is difficult to examine cell. The diaphragm must consist of a light element,
directly the relationship between the observed structures chemically inert, of high hardness, and amorphous. The
and catalytic behavior, which requires instead observation development of a diaphragm is therefore a pressing need.
of the catalyst under reaction gas conditions. The present paper turns attention to the use of boron
The environmental-cell (E-cell) TEM system developed carbon oxynitride films in the diaphragm of the E-cell TEM.
by Ueda et al. includes a specimen holder with a small The films for the diaphragm were prepared using plasma
enhanced chemical vapor deposition (p-CVD) with
TAC film Cu grid for precursor of trimethyl borate (B(OCH 3 ) 3 : TMB) and
Environmental cell mixture gases of nitrogen (N 2 ) and argon (Ar). This
method has the advantage of greater film step coverage than
other coating methods. The prepared films on a Cu mesh
(1) (2) (3) for a specimen holder of TEM were characterized by
Slide glass scanning electron microscopy (SEM) and X-ray
photoelectron spectroscopy (XPS).
(4) (5)
2. Experimental Procedure
Deposited film Etching Figure 1 shows the procedure for diaphragm formation.
First, a triacetyl cellulose (TAC) film was formed on a slide
Fig.1 Procedure for diaphragm formation. glass. The TAC film peeled off from slide glass was put on
a Cu mesh (diameter I 3.5 mm, thickness 1 mm, hole
Chamber
diameter I 100 Pm). A boron carbon oxynitride film was
Source gas deposited on it by p-CVD. Finally, the TAC film was
and carrier gas etched with acetone.
Figure 2 shows a schematic diagram of p-CVD method
experimental equipment. The gas canister and the chamber
were connected with stainless pipes. The source gas and
carrier gas were adjusted using a flow meter between
chamber and gas canisters. The electrodes were isolated
Fig. 2 Schematic diagram of p-CVD method. with glass. The inter electrode distance was 13 mm.
The boron carbon oxynitride films were deposited on
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* Kinki University, Osaka, Japan Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
** Nagoya University, Nagoya, Japan Japan
322
Development of amorphous boron carbon oxynitride film for transmission electron microscope with environmental-
cell system
B-N
Intensity (arb. u.) B 1s B-C
(a) TMB
(b) TMB + N2
(c) TMB + N2 + Ar
B-O
(c)
(b)
(a)
C-C (100 ml/min.) added to the gas flow for sample (b). In the B
(a) TMB 1s spectrum of sample (a), main peak was 189.6 eV
(b) TMB + N2
corresponding to B-C bonding and the spectrum did not
(c) TMB + N2 + Ar
C-N
include a peak at 192-193 eV corresponding to B-O
bonding [5, 6]. The C 1s spectrum of sample (a) included
(a) peaks at 288.5 eV and 283 eV corresponding to C-O and C-
(b)
B bonding, respectively. The atomic ratio of sample (a) was
(c)
about B:C = 1:2. TMB is composed of the chemical
305 300 295 290 285 280 bonding of B-(O-CH 3 ) 3 and the atomic ratio of TMB is
B:C = 1:3. It is therefore thought that oxygen desorption in
Binding energy (eV) the TMB gas was promoted by the generation of TMB
plasma and that film with B-C bonding was deposited by
N 1s
C-N
B-N
recombination of B-C or displacement of TMB.
In the B 1s spectrum of sample (b), main peak was
Intensity (arb. u.)
323
Transactions of JWRI, Vol.39 (2010), No. 2
plasma enhanced CVD with TMB and were characterized [2] M. Haruta: Catalysis Today, 36 (1997), pp.153.
by XPS. The films were composed of B-C, B-N and C-N [3] K. Ueda et al.: Sufrace and Interface Analysis, 40 (2008),
bonding and inclusion of Ar in the source gas enhanced the pp.1725.
dissociation of TMB and the reaction of boron, carbon and [4] Z. F. Zhou et al. : Surface & Coatings Technology, 128-129
(2000), pp.334.
nitrogen. The boron carbon oxynitride films were
[5] M.A Mannan et al.: American Journal of Applied Science, 5
successfully applied to a Cu mesh with holes of I 100 Pm. (2007), pp.736.
[6] M.C. Polo, E. Martinez, J.Esteve, J.L. Andujar: Diamond and
References Related Materials, 7 (1998), pp.376.
[1]M. Haruta: Journal of the Surface Science Society of Japan, 26, [7] Pi-Chuen Tsai: Surface & Coatings Technology, 201 (2007),
No. 10 (2005), pp.578 (in Japanese). pp.5108.
324
Transactions of JWRI, Vol.39 (2010), No. 2
KOBAYASHI Akira *, KURODA Toshio *, KIMURA Hisamichi **, and INOUE Akihisa ***
KEY WORDS: (Metallic glass) (Gas tunnel plasma spraying) (Ni-based metallic glass coating) (Microstructure)
ᴾ ᴾ ᴾ ᴾ ᴾ ᴾ ᴾ ᴾ ᴾ ᴾ (Scanning electron microscope (SEM)) (X- ray diffractmeter (XRD))
325
Properties of metallic glass coatings sprayed by gas tunnel type plasma spraying
The phase constituents of metallic glass coating were amorphous phase (Phase center is about 42 degree) was
identified by using a X-ray diffractmeter (XRD) system observed in this pattern. But the XRD pattern has some
with CuKDradiation source at voltage of 40 kV and current crystalline peaks corresponding to some other compounds
of 40 mA. Obtained results were also discussed and of Ni, Nb, Zr, etc. shown in Fig. 3. Ni crystalline peak was
compared to Fe-based and Zr-based metallic glass coatings. recognized as shown in this figure. This means that the
Vickers microhardness measurement was made on the powder was not a perfectly amorphous phase and it
coating cross section by using a load of 100g. Indentation contains some metal or oxidation phase.
parameters were set as 20s loading time. These crystalline peaks near 40 degree were suppressed
Regarding the Ni-based metal glass powder used in this and disappeared on the metallic glass coating at 300A. The
study, the size of powder was 10-25Pm, and it was broad amorphous phase was clearly observed in the pattern
spherical type. (Average diameter is 18Pm.ὸ which is shown in Fig. 3(a). There were some new peaks
related to Ni, Nb, Zr, etc. at different diffraction angles of
the XRD pattern. Two crystalline peaks near 30degree and
3. Results 50degree were recognized as the peaks from Zr oxide phase.
Microstructure of the metallic glass coating When the metallic glass powder is injected into the plasma
Figure 2 shows the SEM micrographs of the cross- jet, particles under high temperature conditions are heated
section of Ni-based metallic glass coatings sprayed at and may be simultaneously decomposed. So, there is the
different plasma spraying current of I=100A and I=300A, large possibility of the crystalline peaks, and peaks of
at L=40mm.
There are some large pores in the Ni-based metallic
glass coating at a low current of 100A in Fig.ᴾ 2(a). The 㻟㻡㻜
coating thickness increased with increase in the plasma
㻟㻜㻜
current and it was more than 200 ȝm at 300A which is
shown in Fig.ᴾ 2(b). In this case, the coating was much 㻞㻡㻜
denser on the cross section of surface side and it was rather
㻞㻜㻜
lower porosity than that at 100A as shown in Fig. 3(a). The
ᙉᗘ㼇㼍㻚㼡㼉
㻡㻜
㻜
20
㻞㻜 30
㻟㻜 40
㻠㻜 50
㻡㻜 60
㻢㻜 70
㻣㻜 80
㻤㻜
Diffraction angle, 2ș (degree)
ᅇᢡゅ㻞䃗㼇ᗘ㼉
Coating
ίa) Coating at 300A
SUS304
(a) 100A
㻣㻜㻜
(b) 300A
㻢㻜㻜
(a.u.)
ᙉᗘ䠄㼍㻚㼡䠅 (a.u.)
㻡㻜㻜
Intensity
Intensity
㻠㻜㻜
Coating 㻟㻜㻜
㻞㻜㻜
㻝㻜㻜
SUS304
㻜
㻞㻜 㻟㻜 㻠㻜 㻡㻜 㻢㻜 㻣㻜 㻤㻜
20 30 40 50 60 70 80
ᅇᢡゅ㻞䃗
Fig. 2 SEM micrographs of the cross-section of the Ni Diffraction angle, 2ș (degree)
based metallic glass coating sprayed at 100A and
300A, on the 16 times traversed substrate. ίb) Powder
Crystallie structure of the metallic glass coating
The XRD pattern from the surface of the metallic glass Fig. 3 XRD patterns of the Ni-based metallic glass coating
coating at 300A which is shown in Fig. 2(b) is shown in Fig. sprayed on stainless-steel substrate (a) and Ni-based
3(a). For comparison, the XRD pattern of Ni-based metallic metallic glass powder (b).
glass powder used is shown in Fig. 3(b). The broad
326
Transactions of JWRI, Vol.39 (2010), No. 2
other oxidized materials, but there were no peaks from Ni, References
Nb, Zr detected in this XRD pattern. [1] A. Inoue, B.L. Shen and H.M. Kimura: J. Metastable
Nanocrys. Mater., 20-21 (2004), pp.3-12.
[2] Y. Arata, A. Kobayashi and Y.Habara: J. Appl. Phys., 62-12,
4. Conclusions (1987), pp.4884-4889.
The Ni based metallic glass sprayed coating in thickness [3] A. Kobayashi: Surface and Coating Technology, 90 (1990),
of more than 200 ȝm with Hv 100 = 700, was obtained at pp.197-202.
300A by gas tunnel type plasma spraying. The coating [4] A. Kobayashi, T. Kitamura: Vacuum, 59 (2000), pp.194-202.
thickness depends on the spraying parameters. Ni-based [5] A. Kobayashi: Weld.International, 4 (1990), pp.276-282.
metallic glass particles had some decomposition of [6] Y. Arata and A. Kobayashi: J. Appl. Phys., 59 (1986),
amorphous phase and/or oxidation (ZrO 2 ) during pp.3038-3044.
[7] A. Kobayashi, S. Yano, H. Kimura and A. Inoue: Surface
deposition at high plasma current above 350A.
and Coating Technology, 202 (2008), pp.2513-2518.
[8] A. Kobayashi, S. Yano, H. Kimura and A. Inoue: Frontier of
Applied Plasma Technology, 1 (2008), pp.47-50.
327
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Zirconia) (TBC) (Inter-granular fracture) (Columnar grain) (Crack) (Crack propagation)
㻌 㻌 㻌 㻌 㻌 㻌 㻌 㻌 㻌 㻌 (Fracture) (Stress intensity factor) (Energy release rate) (Gas tunnel type plasma spraying)
328
Inter-granular cracking of a splat of zirconia coating fabricated by plasma spraying technique
Crack tip
Crack tip
V V
B C
A D (1) (3)
(2)
V V (4) (5)
V V
(6) (7)
(8) (9)
(10)
V
Fig. 4 Model of a splat of zirconia coating with V
honeycomb-like crystal grains under
equi-biaxial tension. Fig. 6 An example of numerically simulated crack path in
a splat of zirconia coating. The red line ABCD is the
assumed initial crack and the crack is spreading
in the order of (1), (2), (3), ···.
4) The size of the model is sufficiently large comparing
with the cracks.
5) According to the maximum tangential stress criterion energy release rates of four directions are calculated using
[6], the crack is running in the direction T íZKHQ the method proposed by Nuismer [7]. An example of
K II > 0 and T = +60° when K II < 0, where K II is the numerically simulated crack paths is shown in Fig. 6 in
mode II stress intensity factor. (Regarding the detail, which the red line ABCD indicates the assumed initial
refer to [3].) crack and the blue lines are the predicted crack paths. The
Here, instead of the above assumption 5), the following crack is simulated to be propagating in the order of (1), (2),
postulation is adopted. (3), ··· without the branch (2) extending any more. Some
6) Crack propagation occurs by the length a to the other numerical simulations suggested that a crack is
direction in which the energy release rate is the largest extending only at one of its tips showing regular zigzag
among four directions shown in Fig. 5, where a is the pattern, and at another crack tip, it extends only in one
side length of the hexagon of the crystal grain. The branch or never extends.
crack does not extend in multiple directions
simultaneously. 4. Conclusions
The stress intensity factors at the tips of the inter-granular Based on the inter-granular crack propagation observed
crack are calculated by the method shown in [3] and the in a splat of the zirconia coating fabricated by the plasma
329
Transactions of JWRI, Vol.39 (2010), No. 2
spraying technique, numerical simulations of crack [2] R. El-Sheikhy and A. Kobayashi: Transactions of JWRI,
propagation were conducted using a simplified model of Osaka University, 37-1 (2008), pp.69-72.
honeycomb-like crystal grains based on the energy release [3] A. Kobayashi, R. El-Sheikhy and K. Fujimoto: Appl. Plasma
rate. As a result, the regular zigzag crack propagation has Sci., 17-2 (2009), pp.117-124 (in Japanese).
[4] Y. Arata, A. Kobayashi and Y. Habara: Journal of High
been simulated.
Temperature Society, 13-3 (1987), pp.116-124 (in Japanese).
This work was performed under the Cooperative Research
[5] A. Kobayashi, Y. Habara and Y. Arata: Journal of High
Program of Joining and Welding Research Institute, Osaka
Temperature Society, 18-2 (1992), pp.25-32 (in Japanese).
University, Japan.
[6] F. Erdogan and G. C. Sih: Journal of Basic Eng., 85 (1963),
pp.519-527.
References [7] R. J. Nuismer: International Journal of Fracture, 11 (1975),
[1] R. El-Sheikhy and A. Kobayashi: Transactions of JWRI, pp.245-250.
Osaka University, 36-2 (2007), pp.87-94.
330
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Plasma spray) (Tungsten) (Ferritic/martensitic steel) (Coating) (Thermal property) (Heat flux
test) (Fusion device) (First wall) (Divertor) (High heat flux)
2. Experimental
W coated material has been produced by APS and VPS
to evaluate thermal behavior of APS-W and VPS-W. The
substrate material is reduced-activation ferritic/martensitic
steel (RAF/M) F82H (Fe-8Cr-2W) [2]. Size of the substrate
material was 20 mm x 20 mm x 2.6 mmt. A thickness of W
is 1mm. Temperatures of the substrates during the APS
and VPS were 150 oC and 600 oC, respectively. In addition,
mock-ups were made by brazing the tiles (VPS-W/F82H,
APS-W/F82H) on oxygen free high purity copper (OFHC) Fig. 1 Photograph of APS-W/F82H/OFHC mock-up
† Received on 30 September 2010 ***** JAEA, Ibaraki, Japan
* RIAM, Kyushu University, Fukuoka, Japan Transactions of JWRI is published by Joining and Welding
** JWRI, Osaka University, Osaka, Japan Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
*** College of Engineering, Ibaraki University, Ibaraki, Japan Japan
**** NIFS, Gifu, Japan
331
Tungsten coatings on reduced-activation ferritic/martensitic steel by plasma spray technique
Thermal behavior of Tungsten coatings
Figure 2 and 3 show heat flux dependence of plateau strongly dependent on the fabrication process. Therefore,
temperatures measured at the surface, T1 and T2 for VPS- we estimated the thermal conductivity of the present case as
W/F82H/OFHC and APS-W/F82H/OFHC, respectively. It follows. The heat flux q through a material layer at steady
can be seen that the temperatures increased continuously state for plane geometries is given by
with increasing heat flux. Surface temperature of the VPS-
W/F82H/OFHC is always lower than that of the APS-
'T
W/F82H/OFHC; for example, the surface temperatures are q k
about 700 oC and 1200 oC at the heat flux of 3.4 MW/m2, 'x
respectively. In the case of steady state, temperature
increase is inversely proportional to the thermal where 'x is the thickness of material layer of PS-W and
conductivity. The thermal conductivity of plasma spray W F82H, 'T the temperature difference corresponding to
depends strongly on its texture structure and residual
porosity. Cross sectional observation of the APS-W showed the 'x and k thermal conductivity [6]. Estimated thermal
that pores partially existed between W particles. This is one conductivities of APS-W/F82H is 10 W/mk, which is about
of the reasons for the high temperature increase of the W 5.6 % of normal W and 37 % of F82H at RT. On the other
surface of APS-W. hand, estimated thermal conductivities of VPS-W/F82H is
77 W/mk, which is about 43 % of normal W and 285 %
㻝㻠㻜㻜
㻌
of F82H. It is considered that the thermal conductive of
㻝㻞㻜㻜
㼂㻼㻿㻙㼃㻛㻲㻤㻞㻴㻛㻻㻲㻴㻯
㻌㻿㼡㼞㼒㼍㼏㼑
VPS-W/F82H is good enough taking into account thermal
conductivity of F82H.
㼀㼑㼙㼜㼑㼞㼍㼠㼡㼞㼑㻌㻔㼛㻯㻕
㻌㼀㻝
㻝㻜㻜㻜 㻌㼀㻞
The thermal fatigue test up to 100 cycles (3.2 MW/m2,
㻤㻜㻜
40 s ON:40 s OFF) for APS-W/F82H/OFHC and VPS-
㻌
㻢㻜㻜
W/F82H/OFHC showed that temperatures of surface, T1
㻠㻜㻜 and T2 did not change. The surface morphology also did
㻞㻜㻜 not change. In addition, no cracks and exfoliation were
㻜 observed. These results indicate that no failure occurred at
㻜㻚㻜 㻜㻚㻡 㻝㻚㻜 㻝㻚㻡 㻞㻚㻜 㻞㻚㻡 㻟㻚㻜 㻟㻚㻡 㻠㻚㻜
㻴㼑㼍㼠㻌㻲㼘㼡㼤㻌㻔㻹㼃㻛㼙㻞㻕 the interface or in the W coating during cyclic heat load.
Fig. 2 Thermal response of VPS-W /F82H/OFHC 4. Conclusions
㻌 Two types of plasma spray tungsten coatings on
㻝㻠㻜㻜
㻭㻼㻿㻙㼃㻛㻲㻤㻞㻴㻛㻻㻲㻴㻯 ferritic/martensitic steel F82H made by vacuum plasma
㻝㻞㻜㻜
㻌㻿㼡㼞㼒㼍㼏㼑 spray technique (VPS) and atmospheric plasma spray
technique (APS) have been examined to evaluate their
㼀㼑㼙㼜㼑㼞㼍㼠㼡㼞㼑㻌㻔㼛㻯㻕
㻌㼀㻝
㻝㻜㻜㻜 㻌㼀㻞
㻢㻜㻜
332
Transactions of JWRI, Vol.39 (2010), No. 2
†
Behavior of arc plasma for thermal barrier coating preparation
Plasma plume length determined from light intensity
KEY WORDS: (Plasma jet) (Image analysis) (Light intensity) (Gas tunnel type plasma spraying) (Thermal
barrier coatings) (Atmospheric pressure)
1. Introduction
Thermal Barrier Coatings (TBCs) have been
synthesized with plasma aided techniques such as the
thermal plasma spraying method[1-4]. In plasma sprayed
TBCs, flow and thermal characteristics of spraying plasma
affect the performance of TBC materials. Therefore, it is
important to obtain a better understanding of the behavior
of arc plasma during synthesizing the materials.
From this view point, light intensity analysis was
conducted to consider the behavior of the plasma flow in
the preparation of TBC materials by plasma spraying
techniques. The effects of powder feed rate on plasma
plume length were mainly examined in the present study.
The spraying plasma was observed with a digital single-lens
reflex camera. Contours of light intensity in the plasma are
visually determined from RAW image files of the camera
with the software for analyzing the RAW format. Simple
consideration of thermal characteristics was done on the
basis of the obtained results. Fig.1 Image of plasma jet and contour plot of light
intensity (I =300A, without powder).
2. Experimental Procedure
Plasma jet was generated under atmospheric pressure by was set to certain values that were practically used for the
using a gas tunnel type plasma spraying torch [5]. Working experimental studies of plasma sprayed coatings in the case
gas is argon, with a flow rate set to 200 l/min. Powders for of each powder. These were 200A, 300A and 400A in the
TBCs were fed into the plasma plume with carrier gas. case of TiO 2 and HA and 200A and 300A in the case of
Argon was also used as the carrier. The flow rate was ZrO 2 and Al 2 O 3 . The behavior of spraying plasma was
specified to about 5 l/min. In this study, TiO 2 (titanium observed under the same conditions of shutter speed and
dioxide), ZrO 2 (Zirconium dioxide) and Al 2 O 3 (Aluminum exposure time with a digital single-lens reflex camera
oxide) were employed as injected powders. In addition to (Nikon D2x, 12.4Mpixels) through the viewing window
them, HA(hydroxyapatite) was also examined. Feed rate of that can observe the plasma jet flowing out from the torch
each powder is shown in Table 1. The plasma arc current I exit. RAW data format, which records data directly from
the image sensor of the camera and has precise information
Table 1 Feed rate of each powder.
on the light intensity, was employed as image format.
Digital data on light intensity in the plasma are determined
Feed rate (g/min) from the RAW image files with the software for analyzing
Powder
Case 1 Case 2 Case 3 Case 4 Case 5 the RAW format [6]. Digital data on light intensity were
averaged over 10 RAW data sets as illustrated in upper
TiO2 0 <5 ~10 ~15 photograph of Fig. 1. Lower figure of Fig. 1 indicates
HA 0 <5 ~10 ~15 contour plot of light intensity in the domain bounded by the
rectangle-shaped solid line in the upper image. Since the
ZrO2 0 ~5 ~7 ~10 ~15 image sensor of the camera has a Bayer filter, whose pattern
Al2O3 0 ~5 ~7 ~10 ~15 consists of red, blue and two green filters, data on red, blue
and green can be captured. Averaged light intensity
†ᴾ Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
*ᴾ Department of Mechanical Engineering and Intelligent Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
Systems, Tohoku Gakuin University, Tagajo, Japan Japan
**ᴾ JWRI, Osaka University, Osaka, Japan
333
Behavior of arc plasma for thermal barrier coating preparation
Plasma plume length determined from light intensity
between two green filters was mainly used in the present Figure 3 indicates variation of the length 'x with feed
analysis. The length 'x between the torch exit and the rate of powder in the case of ZrO 2 and Al 2 O 3 . Plasma arc
boundary of specified threshold intensity was determined current I was set to 200A and 300A in these cases. 'x is
from the light intensity contours. normalized with respect to ('x) 300A,NP that stands for 'x in
the case of plasma torch current I of 300A without powder.
3. Results and Discussion The results of ZrO 2 and Al 2 O 3 are indicated by squares and
Figure 2 shows variation of the length 'x with feed rate circles, respectively. 'x shows a tendency to shorten due to
of powder in the case of TiO 2 and HA. 'x is normalized the injection of powders in both cases. It also decreases
with respect to ('x) 400A,NP that denotes 'x for a plasma with increases in powder feed rate. These results are
torch current I of 400A without powder. In the case of HA thought to be caused by the somewhat high rate of energy
represented by circles, 'x shows a tendency to decrease by given to powders since plasma arc current is relatively low.
the injection of powder for I =200A and 300A. The In the present experiment, different results were obtained
influence of powder addition on the plume length is not according to the kinds of feeding powders. Therefore, it is
obvious for I =400A. It is thought that the effect of powder necessary to consider not only powder properties but also
injection on the plasma plume becomes less due to higher properties of feeding powders for plasma as working fluid.
input to the plasma generation. On the other hand, the
length is not influenced much by feeding of TiO 2 powder as 4. Conclusions
denoted by squares. This is partly because the flow rate of To obtain a better understanding of the behavior of arc
the plasma jet increases due to the addition of carrier gas. plasma in the preparation of TBC materials by the plasma
Moreover, there does not appear to be a clear effect of spraying techniques, light intensity analysis was conducted
powder feed rate on the length. by utilizing RAW image files. The results obtained here are
summarized as follows.
(1) Different results of the influence of powder injection on
plasma plume length were obtained according to kinds
of injected powders.
(2) The plasma plume length is not obviously affected by
powder injection in the case of higher plasma arc
current.
(3) The plume length shows a tendency to decrease with
increases in feeding rate of ZrO 2 and Al 2 O 3 powders in
the case of relatively lower plasma torch current.
5. Acknowledgement
The authors would like to thank Dr. Yugeswaran
Subramanium for his help in operation of the experiment.
This work was performed under the Cooperative Research
Program of Institute for Joining and Welding Research
Institute, Osaka University.
334
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Titanium oxide) (Photo-catalysis) (Thermal plasma) (Thermal spray) (CVD) (Dye sensitized
solar cell)
2. Experimental Procedure
In this study, two types of SPD equipments, which were a) Straight plasma jet type
straight plasma jet assisted SPD equipment and vortex
plasma jet assisted SPD equipment, were used. These SPD Substrate
equipments consisted of plasma torch, DC power supplying Coating
Plasma jet
system, micro tube pump (feedstock supplying system) and Cathode Anode
working gas supplying system as shown in Fig. 1. Table 1
shows deposition conditions. The plasma torch has water 㸫
cooled electrodes. The anode is made from copper, has
feedstock feeding port at the head and has a constrictor
which is 6 mm in diameter. A cylindrical cathode made +
Airbrush
from tungsten has a diameter of 5 mm. Air was used as
working gas. Mass flow rate of the gas was fixed at 50
l/min. As the feedstock for titanium oxide film deposition, Micro
tube
ethanol diluted titanium tetra iso butoxide (Ti(OC 4 H 9 ) 4 ) DC pump Feed
Air
was used. Substrates were 15mm x 15mm x 1mmt 304 power -stock
stainless steel plates with #400 sand paper polished surface.
The spray distance (the distance between the nozzle outlet b) Vortex plasma jet type
of the plasma torch and the surface of the substrate) was
fixed at 100 mm. The deposition time was 7 min. The Fig. 1 Schematic diagrams of the SPD equipments.
discharge power was 10A/ 100V. In the case of vortex
plasma jet assisted SPD equipment, a commercial airbrush
was used as feedstock spraying equipment.
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* Ashikaga Institute of Technology, Tochigi, Japan Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
** JWRI, Welding University, Osaka, Japan Japan
335
Titanium oxide film deposition by vortex air thermal plasma assisted spray pyrolysis deposition
336
Transactions of JWRI, Vol.39 (2010), No. 2
technique was found to have high potential for high rate and References
low cost functional oxide film deposition. [1] T. Nagamoto, M. Endo, O. Omoto: Jpn. J. Appl. Phys. 18, 6
(1979), pp.1103-1109.
[2] H. Ishizawa: Am. Ceram. Soc. Bull. 65 (1986), pp.1399-1402.
[3] S. Shirakata, H. Kubo, C. Hamaguchi, S. Isomura: Jpn. J.
Appl. Phys. Pt. 2, Letters 36, 10B (1997), L1394-L1396.
337
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Electron density) (Electron temperature) (Frequency shift probe) (Plasma monitoring) (Sheath
effect) (Sheath width) (Resonance frequency)
on finite difference time domain (FDTD) simulation, and it 1.1 Slit 5mm
Frequency [-]
338
Investigations on frequency shift probes for monitoring of electron conditions in nano-materials processing plasmas
339
Transactions of JWRI, Vol.39 (2010), No. 2
340
Dissimilar laser brazing of single crystal diamond and tungsten carbide
99.9 2
Single crystal diamond 1
lnln(1-F)-1
Ag-Cu-Ti Braze 10 -2
-3
1 -4
WC-Co -5
ȝP m = 2.624 -6
0.1
ı0 =130.4MPa -7
Fig.2 Cross section of a single crystal ȝ 03D -8
diamond / braze / WC - Co interface. 0.01 -9
1 10 100 1000
Shear stress (MPa)
a) SEM Image b) Ag c) C Fig. 5 The Weibull distributions of shear strength test.
Single crystal
diamond distribution of Ti was located at the interface. The
distribution of Cu is similar to that of Ti, but its peak
Ag-Cu-Ti shifted to the braze side and the concentration of Cu in the
Braze braze away from the interface was high. This shows the
ȝP existence of interface layers of C-Ti and Cu-Ti. These
d) Cu e) Ti f) Line analyses results strongly suggest the formation of reacted layers like
Min Max TiC and Cu-Ti like Cu 3 Ti at the interfaces of single crystal
&ĺ diamond / Ag-Cu-Ti braze, respectively [5, 6].
Non-destructive test by scanning acoustic microscopy
ĸ7L
was done to evaluate how the melted braze filler spread at
the interface between single crystal diamond and WC-Co
ĸ&X plate. Figure 4 (a) shows appearance of the specimen and
(b) shows the image of scanning acoustic microscopy at the
Fig. 3 Map analysis of a single crystal diamond / interface of the single crystal diamond / Ag-Cu-Ti braze /
braze interface. WC-Co. The black area in the center of (b) is Ag-Cu-Ti
braze which was melted at the joint interface and partly
a) b) spread out on the interface onto the WC-Co plate. There
were no big voids in the black area where melted braze
existed at the joint interface. So the wettability of these
Single crystal
Ag-Cu-Ti Braze braze fillers and single crystal diamond or WC-Co seemed
diamond
to be good.
Figure 5 shows the Weibull distribution of shear
strength test of single crystal diamond / Ag-Cu-Ti braze.
2mm
WC-Co WC-Co In the case of all tests, fracture occurred at the single crystal
diamond side of the specimen. Average shear strength was
Fig. 4 a) Appearance of specimen and b) interfacial 115.9MPa. This suggests that the fracture occurred at the
observation using scanning acoustic microscopy. base material of single crystal diamond and that the
interface of brazing joint has the strength enough to connect
voids were observed. Also the interface between WC-Co single crystal diamond and WC-Co.
and Ag-Cu-Ti braze is smooth.
Figure 3 shows the result of map analysis of a single 4. Conclusions
crystal diamond / Ag-Cu-Ti braze interface. Figure 3(a) The conclusive remarks of this study are summarized as
shows enlargement of the single crystal diamond / Ag - Cu - follows.
Ti braze interface in Fig. 2. Figure 3(b) shows the (1) At the interface of single crystal diamond / Ag-Cu-Ti
distribution of Ag near the interface in the same area of Fig. braze, a concentration of Ti and C was observed. It was
3(a). Fig. 3(c), (d) and (e) show the distributions of C, Cu presumed that the reacted phase existed at the interface
and Ti, respectively. From these distributions of elements, between single crystal diamond and braze.
the concentration of Ti near the interface was observed and (2) No large voids or cracks in the braze area existed from
its thickness was about 2ȝm. non-destructive observation of the interface using
Figure 3(f) shows line analyses of C, Ti and Cu in the scanning acoustic microscopy.
same area of (b) to (e). The distribution of C is decreased (3) Fracture occurred at single crystal diamond side of each
sharply at the interface. Meanwhile, the peak of the specimen after a shear strength test. This suggests that
341
Transactions of JWRI, Vol.39 (2010), No. 2
342
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Laser joining) (Different material) (Insert material) (Diode laser) (Aluminum) (Plastics)
ᴾ (Elastomer)
1. Introduction
Table 1 Chemical treatments for 1050P aluminum
Metals and plastics have been widely used in industrial substrate after mirror buffing.
application such as automobiles, aircraft and electronic Alkaline dipping Acid dipping Second alkaline dipping
devices [1]. Joining a metal and a plastic is necessary and
important from a manufacturing viewpoint. Such joining is (Na2CO3(20kg/m3 )+
Na2SiO3(10kg/m3 ) Alkaline dipping Acid dipping
generally performed using adhesive bonds or mechanical
tools, such as bolts and rivets. These joining processes have (325K-30s)
problems in terms of environmental restriction on the Ļ Ļ 䊼
Washing Washing Washing
emission of volatile organic compounds, the difficulties of
䊼 䊼
mass production, and high production costs. Therefore, Nitric acid (60%) Alkaline dipping
another process strongly needs to be developed for joining 䊼 䊼
a metal and a plastic. Washing Washing
In this study, on laser joining between a 1050P
aluminum sheet and a polypropylene resin sheet with a
newly developed insert sheet, the effects of the 1050P
aluminum sheet having its surface shape and chemical state Table 2 Experimental conditions of laser joining.
changed due to a surface treatment such as alkaline dipping Wavelength 808nm
or acid dipping on the joining properties were examined in Focusing distance 100mm
order to clarify the joining mechanism. Spot diameter ȝP
Defocusing distance 12mm
2. Experimental Exposure mode CW
A 1050P aluminum sheet (1050P) (50×25×1tmm) and a Exposure angle 80° for the joining material
polypropylene sheet (PP) (50×25×2tmm) were used as the
Laser power 200 W
joining materials. Chemical treatments shown in Table 1
Joining Speed 5 mm/s
were done for a mirror surface by buffing in order to clarify
the effect of the chemical state of the 1050P surface on the
joining strength. After the chemical treatment, surface
shape measurement and XPS analysis were conducted on
each treated surface in order to examine the chemical state.
The styrene block copolymer (SBC) thermoplastic
elastomer fabricated into a sheet 50 ȝP in thickness was
used as the insert material, and polarity was given by a
carboxyl group or an amine group for these insert materials
in order to improve the joining ability with 1050P.
The insert sheet was placed between the joining materials,
as illustrated schematically in Fig.1, and laser joining was
conducted under the condition shown in Table 2.
After laser irradiation, joining strength was measured by
the shearing test.
343
Joining mechanism between aluminum and polypropylene resin using insert materials by laser irradiation
treatment shown in Table 1 for the 1050P with mirror showing a similar tendency of the shear strength between
buffing. The surface roughness increased with each the -NH 2 group and the -COOH group as shown in Fig.3
chemical treatment, and the surface then became rougher, demonstrate that the hydrogen bond should be generated
however, there was no significant difference in the surface between the hydroxyl groups on the aluminum surface and
roughness and the surface area with each chemical the -NH 2 group or -COOH group of the given insert
treatment. Therefore, an aluminum surface is etched in the materials. By the way, it was not possible to join 1050P and
first alkali treatment, but that surface is hardly etched in the PP, because no reaction was generated between aluminum
next chemical treatment such as the acid treatment or the and the insert materials on using the insert materials without
second alkali treatment. Thus, in spite of the same alkali a denaturant. These results strongly support the above-
treatment for the buffing or acid treatment, the etching mentioned hydrogen bond between the 1050P and the PP.
effect is remarkably different. This result demonstrates that Table 3 Surface roughness (Ra) and surface area (I Surf
the state of the oxide film which is formed after the buffing area) for various chemical treatments of mirror buffed
or acid treatment is different. No Alkaline Acid Second alkaline
Table 4 shows the shear strength of the specimen after reatment dipping dipping dipping
each chemical treatment. The shear strength with the first 6XUIDFHURXJKQHVV5DȝP 0.063 0.219 0.214 0.207
alkali treatment was improved in comparison with that of Surface area (mm2 ) 0.016 1.548 1.540 1.540
the buffing of a surface that was not treated. The shear
Table 4 Shear strength with various chemical treatments
strength with the next acid treatment then improved further
of 1050P aluminum substrate.
than that with the first alkali treatment. However, the shear
strength with the second alkali treatment was lowered No Alkaline Acid Second alkaline
further than that of the buffed specimen. In spite of the fact treatment dipping dipping dipping
that there was little difference in the surface shape with Shear strength (N) 936.7 1023.0 1122.3 847.7
each chemical treatment based on the results shown in
Table 3, the shear strength after each chemical treatment
was different. These results indicate that the chemical state
of the surface aside from the surface shape seems to affect
the joining strength in this process.
Figure 2 shows the XPS Al 3p1/2 narrow spectra at the
surface with buffing and for each treated surface. The peaks
of the metal and the oxidation state were present for all
specimens, the proportion of the peak of the metal state and
oxide state changing with each treatment. This result
suggests that the chemical state at the outermost surface
varies with each chemical treatment.
In this study, the reactivity between the metal and the
insert material allows improvement by giving the insert
material polarity with a carboxyl group. Because the change
in the chemical state of the aluminum surface by chemical
treatment such as alkali or acid significantly affects the
reactivity with the insert materials, acid-base interaction [2]
is suspected to be responsible for the adhesion model Fig. 2 XPS Al3p1/2 narrow spectra of aluminum
between the 1050P and the insert materials. Degeneration substrates after various chemical treatments.
of the insert material was produced by the -NH 2 group
which was the basic functional group in contrast with the -
COOH group, and similar laser welding was then carried
out; the shear strength was measured in order to clarify the
acid-base interaction. The results of shear strength are
shown in Fig. 3. In comparison with the shear strength of
the insert materials denatured by a -COOH group or an -
NH 2 group, because the shear strength after acid treatment
is the highest though the absolute value is different, the
shear strength due to each treatment tended to be similar.
This result indicates that another joining mechanism exists
in addition to the acid-base interaction.
In general, the adhesive bonds for metals having
organic functional groups such as an OH radical, an -NH 2
group, a -COOH group in the molecule are firmly fixed on Fig. 3 Shear strength with various chemical treatments of
the metal surface by means of the interaction force based on 1050P aluminum substrate.
the hydrogen bond between the hydroxyl groups on metal (a) No treatment, (b) Alkaline dipping,
surface and the organic functional groups [2]. The results (c) Acid dipping, (d) Second alkaline dipping
344
Transactions of JWRI, Vol.39 (2010), No. 2
345
Transactions of JWRI, Vol.39 (2010), No. 2
IWAO Toru *, MORI Yusuke *, SAKAI Tadashi *, TAKI Hiroyuki *, SHIMOKURA Takuya *,
TASHIRO Shinichi **, TANAKA Manabu ** and YUMOTO Motoshige *
KEY WORDS: (Numerical calculation) (Radiation) (Self-absorption) (TIG) (Arc) (Metal vapor) (Welding)
346
Modeling of temperature distribution with metal vapour in pulsed TIG including influence of radiative absorption
electromagnetohydrodynamic model. Absorption is which the base current is constant. The concentration near
calculated only within the control volume in which the anode is low. At the time of the base-to-peak current
emission occurs, and depends only on the temperature and transition, the increase in the downward convective flow in
the radial dimension of the control volume (the absorption the arc, driven by the magnetic pinch force, causes the iron
length). In the full self-absorption model case, the vapour to be transported in the direction of the anode. This
calculation domain is divided into a three-dimensional continues during the period in which the peak current is
Cartesian grid. Incident radiation entering a control volume constant.
from all three adjacent control volumes is considered. The When self-absorption is considered, its effects are
flux of incident radiation from each direction is normalized expected to be more pronounced in the low-temperature
to the area of the interface relative to the control volume’s region than in the high-temperature region. Therefore, the
total surface area. With each passage through the control absorption is greatest at locations where a large amount of
volume, the incident radiation intensity decreases because metal vapour is present at low temperatures. This
of self-absorption and increases according to the influence corresponds to the regions in the periphery of the arc near
of the radiative emission from within the control volume. the anode, and it is therefore these regions whose
temperature is increased. The temperature gradient is also
3. Temperature Distribution with Changing Current important; a high gradient favours localized absorption,
Including Influence of Self-absorption since low-temperature absorbing regions are closer to high-
Figure 1 and 2 show the distribution of the arc temperature emitting regions.
temperature and radiation, for the simple model that takes
account of one control volume and directions, and full self- 4. Conclusions
absorption model that takes account of all three adjacent A time-dependent electromagnetohydrodynamic
control volumes and directions. simulation of pulsed TIG arc welding was performed, with
The results show that the iron vapour diffuses upwards the electrodes and arc included in the computational
toward the cathode at the time of the peak-to-base current domain. The distribution of temperature with vapour
transition. Furthermore, it diffuses during the period in generated from the anode was calculated, using
347
Transactions of JWRI, Vol.39 (2010), No. 2
348
Transactions of JWRI, Vol.39 (2010), No. 2
†
Control of wire melting behavior using coaxial hybrid solid wire
Development of pure Ar–MIG welding
KEY WORDS: (MIG arc welding) (Pure Ar) (Shielding gas) (Column of liquid metal) (Wire melting
behavior)(Coaxial hybrid solid wire) (Instability) (Simulation)
1. Introduction
Coaxial hybrid solid (CHS) wires have been developed temperature, specific heat and thermal conductivity. We
to carry out stable MIG welding in pure Ar shielding gas [1, discuss the influence of material properties on wire melting
2]. The characteristics of weld metal are improved using behavior using a wire melting model of CHS wire [4] and
MIG welding in pure Ar shielding gas (Ar-MIG welding) develop the design guide for CHS wire.
because the amount of oxygen in weld metal, which Fig.1
deteriorates the characteristics of weld metal, decreases [3]. 2. Wire melting behavior of CHS wire
However, the Ar-MIG welding is not used because of arc In order to investigate the influence of the material
instability. If we use a conventional solid wire, a column of properties on wire melting behavior, the simulation is
liquid metal (CLM) occurs at the wire tip (Fig. 1). This carried out [4]. First, we investigate the effect of melting
CLM makes the Ar-MIG welding unstable. In order to temperature. We changed the melting temperature at the
stabilize the Ar-MIG welding, shortening of the CLM is inner part. The wire tip shape and temperature distribution
indispensable. are shown in Fig. 3. The result of Inconel-Steel CHS wire
To shorten the CLM, the CHS wires (Fig. 2) have been is shown in Fig. 3(b). The melting temperature at the inner
developed [1, 2]. The composition of the inner part and the part is set higher than the melting temperature of Inconel
outer part are selected so that the amount of melting at the (Fig. 3(a)). The wire tip shows the convex shape. The CLM
inner part may increase more than that of melting at the is made easily because the inner part becomes long. The
outer part. Instead of generating the CLM, a big droplet melting temperature at the inner part is set lower than the
was then obtained (Fig. 2). As a result, stable Ar-MIG melting temperature of Inconel (Fig. 3(c)). The CLM is not
welding was achieved. made because the wire tip shows the concave shape. It is
The melting behavior depends on physical properties of effective to use the low melting temperature materials at
inner materials and outer materials, such as melting the inner part to shorten the CLM.
Fig.3Distance from contact tip L (mm)
㻯㻸㻹 C
㻭㼞㼏
0.6
0.4 +RRS6WHHO (a) (K)
0.2
,QQHU70 .
0.0 1700
0.2 13 15 16 17 18
14
0.4
+RRS6WHHO
0.6
0.6
0.4 +RRS6WHHO
Fig. 1 Welding behavior of Ar-MIG welding using 0.2
(b)
Inconel (TM =1627K)
conventional solid wire. 0.0
0.2 8 9 10 11 12 13
1000
0.4 +RRS6WHHO
㻵㼚㼚㼑㼞㻌㻔㻵㼚㼏㼛㼚㼑㼘㻕 0.6
0.6
0.4 +RRS6WHHO (c)
㻰㼞㼛㼜㼘㼑㼠 0.2 5
0.0
,QQHU70 .
㻜㻚㻢㻌㼙㼙
0.2 6 7 8 9 10 11
㻝㻚㻞㻌㼙㼙
0.4
+RRS6WHHO 300
0.6
349
Control of wire melting behavior using coaxial hybrid solid wire
Development of pure Ar–MIG welding
Distance from contact tip LC (mm) Table 1 Design guide of CHS wire.
0.6
0.4 +RRS6WHHO (a) Melting Thermal
(K) Specific heat
0.2
0.0
Inner (CInconel X2)
㻌㻌 1700 temperature conductivity
0.2 15 16 17 18
0.4
+RRS6WHHO Inner part Low Small Large
0.6
㻌
㻌
0.2 5 6 7 8 9 10 Ti
0.4
+RRS6WHHO 㻌 300
600
Ti Fe
Steel
0.6 400
Cu
Cu 100
Fe
Ti
Fig. 4 Effect of the specific heat at the inner part on the 200
0
Ti
500 1000 1500 2000 500 1000 1500 2000
wire melting behavior (I =280 A). Temperature T (K) Temperature T (K)
Distance from contact tip LC (mm) (a) Specific heat (b) Thermal conductivity
0.6
800 Fig. 6 Material properties of Cu, Steel and Ti.
0.4 +RRS6WHHO K (a) (K)
0.2
0.0
Inner (䃚Inconel X2)
㻌 1700 which are satisfied with design guide. Three materials (Cu,
0.2 8 9 10 11 12 13 Steel and Ti) are used. The melting temperature of Cu,
0.4
+RRS6WHHO 800
K
Steel and Ti are 1356 K, 1773 K and 1958 K, respectively.
0.6
0.6
+RRS6WHHO 800
(b)
Specific heat and thermal conductivity are shown in Fig. 6.
0.4 K 㻌
0.2
Inconel (䃚Inconel )
To shorten the CLM, we selected the Cu for the inner part
0.0
0.2 8 9 10 11 12 13
1000 materials because of following reasons. (i) Melting
0.4
0.6
+RRS6WHHO 800
K 㻌 temperature is low, (ii) Specific heat is small, (iii) Thermal
0.6
800
conductivity is large. The outer part is Ti or Steel. The
+RRS6WHHO 㻌
0.4
0.2
K (c) melting temperatures of both materials are higher than that
Inner (䃚Inconel /2)
0.0 of Cu. The specific heat of steel is larger than that of Ti.
0.2 8 9 10 11 12
0.4
+RRS6WHHO K 㻌 300 Thermal conductivities of both materials are almost same
800
0.6
level. The effect of specific heat is larger than that of
Fig. 5 Effect of the thermal conductivity at the inner part thermal conductivity. Therefore, we select steel for the
on the wire melting behavior (I = 280 A). outer part.
in Fig. 4(b). The specific heat at the inner part was set to In order to investigate wire melting behavior, the
twice the specific heat of Inconel (Fig. 4(a)). The wire tip simulation was carried out. The result is shown in Fig. 7.
shows the convex shape. The CLM is made easily because The wire tip showed concave shape. So a long CLM was
the inner part becomes long. In contrast, the specific heat at not made in this Cu-Steel CHS wire.
the inner part was set in half of the specific heat of Inconel Fig.7
(Fig. 4(c)). The CLM is not made because the wire tip Distance from contact tip Lc (mm)
shows the trapezoidal shape. It is effective to use the small 㻜㻚㻢 (K)
specific heat materials at the inner part to shorten the CLM. 㻜㻚㻠 㻿㼠㼑㼑㼘
1700
Last, we show the effect of thermal conductivity. We 㻜㻚㻞
㻯㼡
changed the thermal conductivity at the inner part. Figure 5 㻜㻚㻜
1000
㻜㻚㻞 㻞㻝 㻞㻞 㻞㻟
shows the results. The result of Inconel-Steel CHS wire is
㻜㻚㻠
shown in Fig. 5(b). The thermal conductivity at the inner 㻜㻚㻢
㻿㼠㼑㼑㼘
300
part was set to twice the thermal conductivity of Inconel
(Fig. 5(a)). The thermal conductivity at the inner part was Fig. 7 Simulation of Cu-Steel CHS wire (I = 280 A).
set in half of the thermal conductivity of Inconel (Fig. 5
(c)).The tip shape shown in Fig. 5(a) is round. However,
the tip of shape shown in Fig. 5(c) is sharp, and the CLM is Steel
350
Transactions of JWRI, Vol.39 (2010), No. 2
We made Cu-Steel CHS wire to confirm the design melting temperature and thermal conductivity.
guide of the CHS wire. The welding behavior is shown in (3)The design guide for CHS wire is developed, and we
Fig. 8(a). The CLM did not occur and stable welding was demonstrate the effectiveness of design guide by the
possible. Fig. 8(b) shows longitudinal section of wire after simulation and the Ar-MIG welding.
the welding. The amount of melting at center the part was
large. References
Fig. 8 [1] T. Nakamura and K. Hiraoka: IIW Doc.212-1080-05, (2005).
4. Conclusions [2] T. Nakamura, K. Hiraoka and S. Zenitani: Science Technology
(1)The melting temperature and the specific heat have a Welding and Joining, 2008, 13 (1), 25-32.
[3] S. Zenitani, T. Nakamura and K. Hiraoka: Quarterly Journal of
large influence on the wire tip shape, the wire melting the JWS, 2007, 25, (1), 187-195.
rate and generation of CLM [4] T. Nakamura and K. Hiraoka: Welding in the Word, 2009, 53,
(2)The thermal conductivity has an influence on the wire 7/8, R158-R165.
tip shape, but the influence is small, compared to the
351
Transactions of JWRI, Vol.39 (2010), No. 2
KUJI Toshiro*
352
Newly developed novel transparent conductor of Mg(OH)2-C compounds
60 Mg(OH0.87C0.13)2
air [6]. Then the variation in resistance somewhat fat with
Mg(OH0.67C0.33)2 time when the film became almost transparent, although
C
㻌
Mg(OH0.67C0.33)2 had excellent optical transmittance for the visible light and
10-2 again excellent low electric resistance of the order of 10-
2
ȍFP,QFRQFOXVLRQZHKDYHVXFFHVVLYHO\GHYHORSHGa new
transparent conductor of Mg(OH) 2 -C except metal oxides.
10-3
0 1k 2k 3k 4k 5k
Time, T ( sec ) References
[1] K. Kondoh, T. Serikawa, K. Kawabata, and T. Yamaguchi,
Scriprta Materialia 57, 489 (2007).
Fig. 4 Resistivity of the samples. [2] T. Nobuki and T. Kuji, J. Adv. Sci. 19 (2007) 11-15.
possible two types of C atoms in the in the Mg(OH) 2 -C [3] M. Abe, and T. Kuji, J. Alloys Comp. 446–447 (2007) 200-
203.
lattice, i.e., substitution or interstitial type. As mentioned [4] T. Honjo, T. Nobuki, M. Chiba and T.Kuji, J. Japan Inst.
above, there are many interstitial sites large enough for C Metals 72 (2008) 780-784.
atoms to stay in the lattice of Mg(OH) 2 . However, if [5] Legkaya T.N., SOVIET JOURNAL OF SUPERHARD
electrically neutral C atoms stay in interstitial sites. MATERIALS 12 (1990) 1-5.
[6] T. Kuji, T. Honjo, M. Chiba, T. Nobuki and J.-C. Crivello, e-J.
Mg(OH) 2 -C should not be semi-conductor. We first Surf. Sci. Nanotech. 6 (2008) 15-16.
hypothesized the case that neutral C atoms stay in (0,0,1/2).
353
Transactions of JWRI, Vol.39 (2010), No. 2
354
Phase transformation at interface using femtosecond laser irradiation
microstructure of the photomodified surface area was crystalline Fe x Si y compounds formed by the laser
observed using a scanning electron microscope (SEM : irradiation. In the phase diagram of a Fe-Si binary system,
Hitachi S2600N). The atomic distribution at the stoichiometric E-FeSi 2 is composed of 68at.% Fe and
photomodified area was also analyzed using electron probe 32at.% Si. The E-FeSi 2 phase easily transforms into D
micro analysis (EPMA, JEOL, JSM-8000). FeSi 2 and Ȝ-FeSi phases by heat treatment at over 1255K.
We confirmed that our amorphous thin film transformed
3. Results and discussions into E-FeSi 2 phase by heat treatment at 1123K for 6 hours,
Figure 1 shows SEM and the mixture of D- FeSi 2 andȜ-FeSi at 1373K. In Fig.
photographs of the 2(a), no peaks corresponding to the E-FeSi 2 phase existed,
surface of Fe–Si thin film presumably due to the decomposition of E-FeSi 2 during the
before (a) and after (b) excess laser exposure.
the femtosecond laser Fig. 2(b) shows XRD patterns of the Fe-Si thin film
irradiation with a pulse before (i) and after the laser irradiation with different
energy of 1.5PJ. The scanning speeds of (ii)250Pm/sec., (iii)450Pm/sec., and
Fe-Si compound has a (iv)600Pm/sec., respectively. The pulse energy of the laser
Fig.1. SEM photos of Fe-Si
thin film before (a) and after band gap of ~ 0.85eV beam was set to 1.5PJ in all cases. When the scanning
(b) the laser irradiation. and high optical speed was set to 250Pm/sec., both the D-FeSi 2 andȜ-FeSi
absorption, therefore, the phases were observed at the photomodified area.
film would sufficiently absorb the incident laser beam. However, when the scanning speed increased to above
When the laser pulses were focused on the film, the focal 450Pm/sec., their peak intensity decreased and the E-FeSi 2
spot of the laser beam was illuminated, and a dispersion of phase appeared instead. The fast laser scanning at above
white particles less than 600Pm/sec. could form almost the single E-FeSi 2
~3Pm in diameter was polycrystalline phase. Therefore, the phase transformation
observed around the strongly depends on the scanning speed and pulse energy
scanning area as in the Fe-Si system.
indicated in this photo. The initial temperature at the focus is important for the
Figure 2(a) indicates ultrafast crystallization using the laser pulses. The scanning
XRD patterns of the speed is one of the important parameters in the
Fe-Si thin film crystallization of Fe-Si thin films to control cooling or
(i)as-deposited, and after heating rate around the focus. In general, when the
the laser exposure with a femtosecond laser beam is focused inside solids, not only
different pulse energy of the generation of shock wave but also heat accumulation
(ii) 0.25PJ, (iii) 0.75PJ, occurs around the focal point, resulting in the formation of
and (iv)1.5PJ, ion diffusion or refractive index change with a circular
respectively.The scanning shape after the laser irradiation. The effect depends on the
speed of the sample was temperature cooling rate: if the cooling rate is low, the
set to 250Pm/sec. The effect tends to occur around the focal area. We have
exposed area was calculated the cooling rate near the focal point in the case
apparently crystallized of the laser irradiation on FeSi 2 thin film by solving the
and two phases ascribed thermal diffusion equation. The rate is calculated to ~1012
to D-FeSi 2 and Ȝ-FeSi K/s, which is much faster than that of the commercial glass
phases were identified in sample (~ 108 K/s), due to its large thermal diffusivity (D th:
the XRD spectra. In FeSi = 0.65) rather than that of glass sample (D th:glass =
-6
addition, the crystallinity 0.46u10 ). Therefore, we can estimate that the focal
Fig. 2. (a)XRD patterns of the
of these phases temperature increases dramatically and relaxes to room
Fe-Si thin film after the laser
irradiation with different pulse improved as the pulse temperature before the next pulse arrives. The phase
energy;(i)as-deposited, energy of the laser transformation completed in a very short time before the
(ii)0.25PJ, (iii)0.75PJ, and pulses increased, heat accumulated.
(iv)1.5PJ, respectively. (b) XRD
patterns of the film after the laser especially in the case of
4. Conclusion
irradiation at different scanning 1.5PJ. The white
speeds;(i)as-deposited, Phase transformation in amorphous Fe-Si film was
particles shown in Fig.
(ii)250Pm/sec., (iii)450Pm/sec., demonstrated using femtosecond laser irradiation. The
and (iv)600Pm/sec., respectively. 1(b) were due to the
formation of D-FeSi 2 , E-FeSi 2 andȜ-FeSi depended on
355
Transactions of JWRI, Vol.39 (2010), No. 2
the focal energy from the laser : the moderate irradiation 5. References
[1] S. Kanehira, J. Si, J. Qiu, K. Fujita, and K. Hirao, Nano Lett.,
tends to form the E-FeSi 2 phase. To control the photon
5(2005), 1591.
energy will open up the various photo-reactions at the [2] S. Kanehira, K. Miura, K. Fujita, K. Hirao, J. Si, N. Shibata,
interface. and Y. Ikuhara, Appl. Phys. Lett., 90(2007), 163110.
[3] S. Kanehira, K. Miura, and K. Hirao, Appl. Phys. Lett., 93
(2008), 023112.
[4] S. Kanehira, M. Eida, M. Sakakura, Y. Shimotsuma, K. Miura,
and K. Hirao, Appl. Phys. A (in printing, 2010).
356
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Ion-exchange) (Borosilicate glass) (Metal nanoclusters) (UV nanosecond laser)
ᴾ ᴾ ᴾ ᴾ ᴾ ᴾ ᴾ ᴾ ᴾ ᴾ (Micro-machining)
357
Laser Micro-machinability of electric field-assisted ion-exchanged glass
Glass surface
Cu (a) Sample D
0 20 40
Fig. 1 Change in penetration depth of copper ions with ion- Penetration depth (Pm)
exchange time. Sample E was prepared by
conventional ion-exchange method in molten salt Fig. 2 Comparison of ion distribution profiles in glasses
bath of 1.0%AgNO3-NaNO3 at 623 K. prepared by (a) FAIE (sample D) and (b)
conventional ion-exchange (sample E).
(a) 10 shot on as-received glass (b) 20 shot on as-received glass (c) 80 shot on Cu-doped glass
(d) 100 shot on Cu-doped glass (e) 120 shot on Cu-doped glass (f) 1 1'
2 2'
3 3'
4' 5' 22.2 Pm
4 4'
4 5 27.4 Pm
5 5'
59.4 Pm
50 Pm
50 Pm
Fig. 3 SEM photographs of laser-ablated square holes formed on as-received and Cu-doped glasses. (a) 10 shot and (b) 20 shot
irradiation of as-received glass; (c) 80 shot, (d) 100 shot and (e) 120 shot irradiation of Cu-doped glass. The cross-sectional
profiles of (a–e) are shown in (f). The laser power was 5.0 J/cm2.
were observed at the processed surface after 10 pulses ripple patterns formed by diffracted laser light were
irradiation on sample R. And the processed hole was observed at hole bottoms. These results suggested that the
seriously damaged even by 20 pulses irradiation (Fig. 3(b)). input energy was effectively absorbed in copper-containing
Many large (-50 Pm) chippings occurred at ablated edges, region and rapid ablation of surface layers occurred.
and irregular-shaped and porous structures appeared at the However, the flat bottom face was drastically damaged by
hole bottom. Although the formation mechanism of these 120 pulses irradiation, as shown in Fig. 3(e). Although the
structures was still unknown, it was assumed that micro-pits sharp ablated edge was maintained, the irregular-shaped
formed by multiple laser irradiation played absorption and porus structure, which was very similar to that observed
centers of laser beams and accelerated the generation of on sample R, appeared at hole bottom. As shown in Fig. 3
larger chippings. On the other hand, such defects or (f), the removed depth by 100 pulses irradiation was 27.4
structures were not observed in 80 and 100 pulses irradiated Pm and this value was close to the penetration depth of
holes on sample D (Fig. 3(c) and (d)). Both of these holes copper ions in sample D (see Fig. 1). It was assumed,
had extremely flat hole bottoms, sharp ablated edges and therefore, that the laser machinability degraded at 120
small heat affected zones around the holes, although faint pulses irradiation because the hole bottom reached a non-
358
Transactions of JWRI, Vol.39 (2010), No. 2
copper-containing region. In addition, the ablation rates was necessary for the high-accuracy fabrication of micro-
calculated from hole depths by 80, 100 pulses irradiations components. Numerical prediction of ion behavior will be a
were 0.278 (0 - 80) and 0.260 Pm/pulse (80 - 100 pulses). prospective method for this purpose. The effects of other
These values indicated that the copper-containing region metal ions in glasses on optical characteristics and laser-
formed by FAIE had constant ablation rate over exchanged machinabilities will be presented in our future work.
area.
References
4. Conclusions [1] K. Tsunetomo, T. Koyama: Opt. Let., 22 (1997) 411.
[2] T. Koyama, K. Tsunetomo, Jpn. J. Appl. Phys., 36 (1997)
In this report, an improvement method of laser-
L244.
machinability of borosilicate glass by electric field-assisted [3] M. Shojiya, H. Koyo, K. Tsunetomo, Proc. SPIE, 5063 (2003)
solid-state ion-exchange was proposed. Experimental 124.
results showed that the ion-exchanged copper-containing [4] M. Rosenbluh, I. Antonov, D. Ianets, Y. Kaganovskii, A. A.
region had the excellent laser machinability. However, the Lipovskii, Opt. Mater., 24 (2003) 401.
processing quality suddenly deteriorated when the removed [5] S. Matsusaka, T. Watanabe, J. Mat. Proc. Tech., 202 (2008)
depth exceeded the penetration depth of copper ions. 514.
Therefore, well-designed and controlled ion distribution [6] R. Oven, J. Appl. Phys., 101, (2008) 113113.
359
Transactions of JWRI, Vol.39 (2010), No. 2
YASUDA Kiyokazu *, SAKINO Yoshihiro **, SHOHJI Ikuo *** , TAKEMOTO Tadashi **
KEY WORDS: (Lead-free solder) (Tensile test) (Strain rate sensitivity) (High strain rate) (Sn-Ag-Cu)
(Ductility) (Rupture) (Stress-strain curve) (Tensile strength) (Yielding point)
2. Experimental
From the fundamental physical point of view, the
mechanical properties of Sn-Ag-Cu solders (Ag content 1.0
- 4.0 vol %, Cu content 0.5-1.0 vol %) were investigated
using a high speed universal tensile test machine in the high
strain rate range (1.610-3s-1 1.6101 s-1). Cylindrical
test rods (diameter: 10mm, gauge length: 50mm) were
provided for the tensile tests according to the testing
method described in JIS Z 3198-2 adopted so far for the 6Q$J&X 6Q$J&X
normal tensile test [4]. For Sn-Ag-Cu based lead-free solder
alloys, mechanical properties were evaluated by a uniaxial Fig. 1 Macroscopic fracture observation of lead-free
tensile test method by using the ultra-dynamic structural solder after high-strain-rate tensile test (specimen
testing system (maximum dynamic load: 1,200 kN, diameter: 10mm, strain rate: 1.6ᶣ101 s-1).
360
Dynamic mechanical behavior of Sn-Ag-Cu lead-free solders by tensile test under high strain rate
4. Conclusions
Tensile tests under high strain rates of 1.6ᶣ100 s-1to 1.6
ᶣ 101 s-1 were performed for the evaluation of the
mechanical properties of several kinds of Sn-Ag-Cu lead-
free solders. All Sn-Ag-Cu alloys showed ductile fractures.
Eutectic composition revealed the highest strength under
the high strain rate conditions. The tensile strength
increased by 200-300 % compared to the value of the static
tensile conditions at a typical strain rate of 1.6ᶣ10-3 s-1. It
was found that the load-elongation curve of each lead-free
solder exhibited a clear elastic deformation under the
highest strain rate of 1.6ᶣ101 s-1. The rupture elongation
did not show any clear dependence on the strain rate.
Acknowledgements
This study was originally started as a research project on
the testing of the mechanical properties of lead-free solders
in the Solder and Smart Joining Division of the Japan
Welding Engineering Society. The authors are grateful for
the support of the concerned division members.
References
Fig. 3 Tensile strength and elongation of lead-free [1] K. Kaminishi: Proc. 9th Symp. on Microjoining and Assembly
solder in high-speed tensile testing Technology in Electronics (2003) , pp.253-256.
[2] I. Shohji, T. Osawa, T. Matsuki, Y. Kariya, K. Yasuda, and T.
composition. Among those, for Sn-3Ag-0.5Cu known as the
Takemoto: Mater. Trans., 49 (2008), pp.1175-1179.
most commonly-used eutectic lead-free solder, the extent of [3] K. Yasuda, I. Shohji, and T. Takemoto: Mater. Science forum,
the necking was remarkable. The area of the fracture part 580-582 (2008), pp.221-224.
was very small. Similarly, another nearly eutectic lead free [4] T. Takemoto: Proc. 6th Electronic Packaging Technology
solder, Sn-3.5Ag-0.7Cu exhibited the sharp cup-and-cone (2005), pp.8-12
361
Transactions of JWRI, Vol.39 (2010), No. 2
Oxide nanosheets and their assemblies for new ceramic joining and
†
smart processing
KEY WORDS: (Oxide nanosheet) (Colloidal suspension) (Layer-by-layer assembly) (Ceramic joining)
(Smart process)
362
Oxide nanosheets and their assemblies for new ceramic joining and smart processing
References
[1] M. Osada and T. Sasaki: J. Mater. Chem. 19, (2009), pp.
2503-2511.
[2] A. K. Geim and K. S. Novoselov: Nature Mater., 6, (2007), pp.
183-191.
[3] T. Sasaki, Y. Ebina, T. Tanaka, M. Harada, M. Watanabe and
G. Decher: Chem. Mater., 13, (2001), pp. 4661-4667.
[4] K. Akatsuka, M. Haga, Y. Ebina, M. Osada, K. Fukuda, T.
Sasaki: ACS Nano, 3, (2009), pp. 1097-1106.
[5] T. Shibata, K. Fukuda, Y. Ebina, T. Kogure and T. Sasaki:
Adv. Mater., 20, (2006), pp. 231-235.
[6] M. Osada, Y. Ebina, K. Takada, and T. Sasaki: Adv. Mater.,
18, (2006), pp. 295-299.
[7] M. Osada, Y. Ebina, H. Funakubo, S. Yokoyama, T. Kiguchi,
K. Takada and T. Sasaki: Adv. Mater., 18, (2006), pp. 1023-
1026.
[8] B-W. Li, M. Osada, T. C. Ozawa, R. Ma, K. Akatsuka, Y.
Fig. 4 Cross-sectional high-resolution TEM image of a 5-
Ebina, H. Funakubo, S. Ueda, K. Kobayashi and T. Sasaki: Jpn.
layer (7.5 nm thick) Ca2Nb3O10 film on a SrRuO3 substrate. J. Appl. Phys., 48, (2009), pp. 09KA15-1-5.
[9] M. Osada, K. Akatsuka, Y. Ebina, K. Ono, H. Funakubo, K.
A clear benefit of these LbL approaches is the interface Takada and T. Sasaki: ACS Nano, 4, (2010), pp.5225-5232.
engineering, which appears to be a key step in the design of [10] M. Osada, Y. Ebina, K. Fukuda, K. Ono, K. Takada, K.
film properties. Physical methods such as vapor deposition Yamaura, E. Takayama-Muromachi and T. Sasaki: Phys. Rev.
and laser ablation are currently the main methods of B, 73, (2006), pp. 153301-1-4.
fabricating oxide films. These techniques, however, usually [11] X-P. Dong, M. Osada, H. Ueda, Y. Ebina, Y. Kotani, K. Ono,
require a complex and difficult deposition process S. Ueda, K. Kobayashi, K. Takada and T. Sasaki: Chem.
involving high-temperature postannealing (>600°C), which Mater., 21, (2009), pp. 4366-4373.
can cause degradation in the film-substrate interface arising [12] M. Osada, M. Itose, Y. Ebina, K. Ono, S. Ueda, K.
Kobayashi and T. Sasaki:Appl. Phys. Lett., 92, (2008), pp.
from both nonstoichiometry and thermal stress. The
253110-1-3.
bottom-up fabrication using oxide nanosheets provides new [13] Z. Liu, R. Ma, M. Osada, N. Iyi, Y. Ebina, K. Takada and T.
opportunities for room-temperature fabrication of oxide Sasaki:J. Am. Chem. Soc., 128, (2006), pp. 4872-4880.
thin films, while eliminating integration problems [14] R. Ma, M. Osada, L. Hu, and T. Sasaki:Chem. Mater., 22,
encountered in current film-growth techniques. Such LbL (2010), in press.
assembly is also expected to have great potential for
advances in new ceramic joining technology.
363
Transactions of JWRI, Vol.39 (2010), No. 2
NUMAKO Chiya *, SATO Kazuyoshi **, ABE Hiroya and OHARA Satoshi ***
KEY WORDS: (Biogenic magnetite) (Radular teeth) (Chiton) (EPMA) (XRD) (XAFS) (Syncrotron radiation)
FP
Acanthopleura japonica
364
Evaluation of magnetite and related iron compounds in the teeth of chiton using X-ray and electron analyses
special material for junction of these parts could not be strength of the teeth supported with other mineral
observed even with high resolution SEM. In addition, components where they act a roll as shock absorber.
they were highly integrated mechanically so that a knife Affinity of iron and calcium phosphate parts also would
crack runs in the tooth without exfoliation of these two work to make a extreme strong structure for the tooth
parts. mechanically. Then, mechanical property of the teeth is
From these observations, it could be considered that much higher than those of other marine creatures at
the hardness of the teeth is due to the magnetite, the intertidal zone.
365
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Nanoparticle) (Aggregation) (Surface treatment) (Dispersion) (Bead milling) (hybrid material)
ᴾ ᴾ ᴾ ᴾ ᴾ ᴾ ᴾ ᴾ ᴾ ᴾ (Organic solvent) (Smart materials) (Powder Technology) (Colloid probe AFM)
366
Characterication and control of nanoparticle dispersion behavior for smart processing in the liquid phase
367
Transactions of JWRI, Vol.39 (2010), No. 2
368
Transactions of JWRI, Vol.39 (2010), No. 2
†
Developments of real-time monitoring method of welding
KEY WORDS: (Digital signal analysis) (Arc voltage) (Welding current) (In-process-monitoring)
(Gas shield arc welding)
2. A developed algorithm
Example scatter plots of transformed data calculated
from blocks of 0.5 second sampled voltage and current data,
as shown in Fig. 1. The data shown in Fig.1 are obtained by
normalizing the reference data. It can be seen that the data
from a reference weld Fig.1(a) are located near the center.
However, the distributions of data shown in Fig.1(b)
deviate from the point of origin and incline to the right.
The analysis method of the proposed system is based on
Mahalanobis distance (MD). The calculation of MD is
described in the field of statistics. To elucidate MD, a
conceptual diagram of MD is shown in Fig. 2. Points A and
Fig. 1 Example scatter plots of transformed data
B are located on the same circle and the Euclidean
calculated from blocks of 0.5 second sampled
distances of two points are equal. However, point B is far
voltage and current data (a) reference and
from the data distribution. Therefore, the shape of the data
(b) comparison.
is considered; this is where MD shows superiority to the
Euclidean distance. With this concept, a set of points
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* Tokushima Prefectural Industrial Technology Center, Japan Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
** JWRI, Osaka University, Osaka, Japan Japan
369
Developments of real-time monitoring method of welding
According to the theory above, a new piece of equipment test weld. Considering arc voltage and welding current are
has been developed. The developed system has the unstable after the welding start, the reference was selected
capability to detect and quantify welding quality in real 4.5 to 5 second data. The sampling rate was about 4000.
time. Quality was calculated every 0.5 second from 5.5 second
after start.
The change of the average voltage is not confirmed, but
the welding quality changes on the butt line. This developed
algorithm is considered to be of benefit for the fault that is
difficult to detect by monitoring only average voltage.
B A
370
Transactions of JWRI, Vol.39 (2010), No. 2
371
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Welding distortion) (Image measurement) (3D measurement) (Stereo imaging technique)
(Iterative substructure method) (Thermal elastic plastic analysis)
(a) Photo taken by left cameraᴾ ᴾ (b) Photo taken by right camera
Fig. 1 Photographs of specimen whose angle of joint plane is 30 degree.
[mm]
(a) Image before weld (b) 1st layer (c) 2nd layer (d) 3rd layer
Fig. 2 Distribution of axial displacement measured by 3D measurement
based on stereo imaging technique.
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* Graduate School of Engineering, Osaka University, Japan Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
** Graduate School of Engineering, Osaka Prefecture Japan
University, Japan
372
Image measurement of welding distortion of pipe joint in two-phase flow separator
r
r
r
VW
OD\HU
r
r
QG UG r
VW QG UG
OD\HU OD\HU OD\HU
r OD\HU OD\HU
7LPHVHF
7LPHVHF
(a) Axial shrinkage (b) Translation of pipe
Fig. 3 Influence of inclination angle of joint plane on welding distortion.
be measured simultaneously with high accuracy. indicates that the influence of the inclination angle on the
To study the influence of the inclination angle of the joint welding deformation in axial shrinkage is small. On the
plane on the welding deformation, four specimens with other hand, the deformation in inclined direction becomes
different inclination angles of welding plane, namely 0o, large when the inclination angle of the weld plane becomes
10o, 20o and 30o are used. The diameter and thickness of large.
pipe are 168.3mm and 3.4mm respectively. The thin base
plate is 400x400x29 mm. Stainless steel SUS304 is used in 3. Thermal elastic plastic FE analysis using Iterative
both pipe and thick plate. The total welding passes and Substructure Method (ISM)
layers are ten and three. The first and second layers have The same models as the experimental specimens are
four quarter circular passes, respectively. The third pass has used for the FE analyses. After the thermal conductive
two half circular passes. Measurement is carried out before analysis using a moving heat source, the Iterative
welding, after each welding pass and after cooling by taking substructure method (ISM) as high-speed 3-dimensional
photographs by two digital cameras as shown in Fig. 1. thermal elastic plastic FE analysis developed by the authors
Additionally, the specimens are sprayed to collate easier. is performed.
Figure 2 shows the axial displacement distribution of 0o Figure 4 show the result of the welding deformation in
model when the each layer is welded. It is seen that the the axial direction in case of the 0o model. From this figure,
axial shrinkage becomes larger when the layer increases. it is found that the axial shrinkage becomes larger when the
Furthermore, in the axial direction, the pipe is mainly layers increase. This tendency is the same as that of the
deformed and the deformation of the thick plate is very experiment. The influence of the inclination angle of joint
small. In this measurement, it is difficult to use the imaging plane on the welding deformation in inclined direction is
technique where the inclination angle to the camera is large also the same as the experimental result.
or where the spray is not carried out such as the welding
line. In these cases, the results are ignored.
The influences of the inclination angle of the welding 4. Conclusions
plane on the axial and inclined direction are summarized in The three-dimensional displacement measurement
Fig. 3. The displacements of the top of the pipe when the system based on the stereo imaging technique is applied to
welding of each layer is finished are plotted. This figure the welding deformation problem of weld joint model of
373
Transactions of JWRI, Vol.39 (2010), No. 2
two phase flow separator and serial computation are the stereo imaging technique based on the digital image
performed by Iterative substructure method as high-speed correlation method and the displacements in axial and
thermal elastic plastic FE analysis. The following inclined direction can be calculated.
conclusions are drawn. (2) When the inclination angle of the welding joint becomes
(1) It is confirmed that the welding deformation distribution lager, welding deformation in inclined direction
of pipe can be measured by measurement system using becomes larger.
374
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Austenitic stainless steel) (Coatings) (Hot hardness) (Intermetallic compound) (Reaction layer)
(Silicide) (Spark plasma sintering) (Surface modification) (Wear resistance)
375
Microstructure of Cr3Si coatings on austenitic stainless steel by spark plasma sintering
6866 +9
+DUGQHVV+91 1
'HQVLW\ˮ NJ P
:HDUORVVPP
+DUGQHVV
'HQVLW\
&RDWLQJWHPSHUDWXUH 7 . Fig. 4 Wear loss of Cr-Si coating layer
the and SUS310S substrate.
Fig. 2 Relationship between hardness and density
of Cr-Si coating layer treated for 10.8 ks at 1073-
1323 K by the SPS process.
50 .
7KLFNHQHVV RIUHDFWLRQOD\HUPP
1000
40 +DUGQHVV+91 1 500
.
30
100
20 50
٨Cr-Si coating layer
٤SUS310S
10
10
400 600 800 1000 1200
7HPSHUDWXUH7 .
0
0 20 40 60 80 100 120
+ROGLQJWLPHWV Fig. 5 Hot hardness test of Cr-Si coating
layer and SUS310S.
Fig. 3 Effect of holding time on the thickness
of reaction layer in samples coated at 1273 K. higher hardness than SUS310S substrate, even at high
temperature in both Cr-Si coating layr and SUS310S
1323 K by the SPS process. As the coating temperature substrate was observed, as shown in Fig. 5. Transition
increased, the densification was developed and the hardness temperatures of Cr-Si coating layer and SUS310S substrate
of these coating layers also increased up to 1150 HV. were 873 K and 723 K, respectively. These results indicate
Also, a reaction layer with the thickness of 40 Pm and that the hardness of 700 HV relatively up to 873 K.
the hardness of 700 HV was formed between the coatings Therefore, it was concluded that the Cr-Si coating layer
and the substrate in the sample treated for 10.8 ks at 1273 K. could be applied to some parts that require wear resistance
Figure 3 shows the relationship between the thickness at high temperature.
of the reaction layer and holding time. The growth of the
reaction layer showed a parabolic relation against holding 4. Conclusions
time. Cr-Si intermetallic compound layers were coated on
Figure 4 shows the wear loss of Cr-Si coating layer and SUS310S substrate by the SPS process to improve its wear
SUS310S substrate. The SUS310S substrate had poor wear and heat resistance. The conclusions of this investigation
resistance. In comparison, Cr-Si coating layer showsᴾ are summarized as follows.
considerably less wear. This confirmed that the wear (1) XRD studies revealed that the coating layer treated for
property of SUS310S was improved by the SPS process. ᴾ 10.8 ks at 1173-1323 K consisted of Cr3Si, Cr5Si3, and
Figure 5 shows a hot hardness result tested at unreacted Cr and coatings treated at 1073-1123 K
temperatures ranging from room temperature to 1073 K for consisted of Cr3Si, Cr5Si3, and unreacted Cr and Si.
Cr-Si coating layer and SUS310S substrate. Cr-Si coating (2) The hardness of the coating layer showed 1150 HV at
layer showed higher hardness than SUS310S substrate, the maximum.
even at high temperature. A gradual decrease in hardness up (3) The reaction layer with a thickness of 40 Pm and a
to a transition temperature, followed by a rapid drop above hardness of 700 HV was formed between the coatings
376
Transactions of JWRI, Vol.39 (2010), No. 2
377
Transactions of JWRI, Vol.39 (2010), No. 2
MORISADA Yoshiaki *, FUJII Hidetoshi **, MIZUNO Tadashi ***, ABE Genryu ***,
NAGAOKA Toru * and FUKUSUMI Masao *
KEY WORDS: (Friction stir processing) (Thermal spray) (Cemented carbide) (Tool steel) (Hardness)
378
Modification of thermally sprayed cemented carbide layer by friction stir processing
and close to that of the monolithic WC. The obtained microhardness. The improvement of the mechanical
microhardness of 2000 HV was nearly the same value as properties of the cemented carbide has been studied by the
that of the sintered WC. It should be mentioned that the decrease in the grain size of the WC particles [5]. For
microhardness of the FSPed zone was higher than that of example, the spark plasma sintered WC-Co with the WC
the sintered cemented carbide with the same chemical particles of 㹼100 nm and the hot pressed WC-Co with the
composition. It is difficult to explain such a large increase WC particles of 169 nm had values of 1887 and 2084 HV,
in the microhardness by only the elimination of the defects respectively [6, 7]. On the other hand, there was no
in the as-sprayed layer. It is considered that the grain effective process to refine the microstructure of the metallic
refinement of the Ni and Co binder and the rearrangement binder in the cemented carbide. Although the grain of the
of the WC particles assisted in the change of the metallic materials could be refined by the recrystallization
As-sprayed WC-CrC-Ni layer FSPed WC-CrC-Ni layer process, it was difficult for the cemented carbides because
of their high deformation resistance. In this study, the
Cr3C 2 plastic flow was successfully induced to the thermally
sprayed cemented carbide layer using the rotating tool made
SEM image
Porosity of WC-Co.
2500
䕦: as-sprayed WC-CrC-Ni layer
WC
䖃: FSPed WC-CrC-Ni layer
䕧: as-sprayed WC-Co layer
䕿: FSPed WC-Co layer
2000
Microhardness / Hv
TEM image
1500
1000
Fig. 1 Microstructural change of the WC-CrC-Ni layer
by the FSP.
500
㻙㻤 㻙㻢 㻙㻠 㻙㻞 㻜 㻞 㻠 㻢 㻤
References
[1] H.S. Ni, X.H. Liu, X.C. Chang, W.L. Hou, W. Liu, and J.Q.
Wang, J. Alloys Compd. 467 (2009) 163.
[2] T.Y. Cho, J.H. Yoon, K.S. Kim, K.O. Song, Y.K. Joo, W.
Fang, and S.H. Zhang, Surf. Coat. Technol. 202 (2008) 5556.
[3] D. Triantafyllidis, L. Li, and F.H. Stott, Surf. Coat. Technol.
201 (2006) 3163.
Fig. 2 TEM images of the as-sprayed and the FSPed [4]S.H. Zhang, T.Y. Cho, J.H. Yoon, M.X. Li, P.W. Shum, and
cemented carbide layers. S.C. Kwon, Mater. Sci. Eng. B 162 (2009) 127.
379
Transactions of JWRI, Vol.39 (2010), No. 2
[5] Z.Z. Fang, X. Wang, T. Ryu, K.S. Hwang, and H.Y. Sohn, Int. [7] C.G. Lin, E. Kny, G.S. Yuan, and B. Djuricic, J. Alloys
J. Refract. Metals and Hard Mater. 27 (2009) 288. Compd. 383 (2004) 98
[6] L.H. Zhu, Q.W. Huang, and H.F. Zhao, J. Mater. Sci. Lett. 22
(2003) 1631.
380
Transactions of JWRI, Vol.39 (2010), No. 2
OKURA Ichiro *
KEY WORDS: (Aluminum alloy) (Ultimate strength) (FSW) (MIG welding) (Joining location) (FEM)
(a) MIG welding (b) FSW The distributions of the residual stress are given in Fig.
2 [1]. For the edge-joined plate, the tensile residual stress of
Fig. 1 Joining locations
V rt V j 0.2 is created in the region of 25 mm from each
3. FEM Analysis edge, and the compressive one of V rc 50V j 0.2 / b 50
The ultimate strength of the edge- and middle-joined
plates in compression is investigated by the elastic-plastic in the inside. Here b is the width of the plate in mm. On
large deflection analysis with FEM. The analyzed model is the other hand, for the middle-joined plate, the tensile
a square plate which is simply supported around the edges.
25
In A6061-T6 alloy, the strength at the joint is reduced to
b-50
the half of that of the parent material. The softening region 2
is 25 mm on each side of the center of the joint, as shown
by the shaped parts in Fig. 1 [1]. b-50 50
The relationship between stress and strain is given by
the following [1]: b-50
For the parent material, 2
25
§ V § V ·
n
Ȫrc Ȫrt
¨H Ȫrc Ȫrt
¨
0.002¨¨ ¸
¸ V d V 0.2 ࠉ
¨
E © V 0.2 ¹ (1) (a) Edge-joined plate (b) Middle-joined plate
¨V
© V 0.2 V ! V 0.2 ࠉ
Fig. 2 Distributions of residual stress
381
Ultimate strength of aluminum alloy plates in compression considering joining locations
residual stress of V rt V j 0.2 is created in the region of 50 5. Ultimate Strength of A5083-O Alloy Plates
mm in the middle, and the compressive one of The relation between V u V p 0.2 and R p for the
V rc 50V j 0.2 / b 50 in the rest of the plate. A5083-O alloy plates is shown in Fig. 5. Here,
For the initial deflection of the plate, the following V p 0.2 V 0.2 . In this analysis, the residual stress is taken
equation is assumed, which is the same as the buckling into account. The influence of the width of plate on the
shape: ultimate strength is small for both the edge- and middle-
w0 b / 150 sin Sx / a sin Sy / b ᴾ (3) joined plates.
where a is the length of the plate in the direction of the ıu/ıp0.2
1.0
compression. b=200mm
0.8
4. Ultimate Strength of A6061-T6 Alloy Plates
The relation between V u V p 0.2 and R p for the 0.6
Residual stress
A6061-T6 alloy plates is shown in Fig. 3, in which the 0.4
residual stress is not introduced. V u is the ultimate 0.2
No residual stress
0.0
and R p are defined as follows: 0.0 0.5 1.0 1.5 2.0
Rp
V p 0.2 ^(b 50) / b`V 0.2 50 b V j 0.2 (4)
ıu/ıp0.2
(a) Edge-joined plate
1.0
1 12(1 P 2 ) V p 0.2
Rp E (5) b=200mm
S 4 E 0.8
where b is in mm, P the Poisson’s ratio, E ( b t ) the 0.6
width-to-thickness ratio, and t the plate thickness. Residual stress
0.4
As shown in Fig. 3(a), in the edge-joined plate,
No residual stress
V u V p 0.2 decreases as the width is smaller. On the other 0.2
Euler's buckling curve
hand, as shown in Fig. 3(b), in the middle-joined plate, 0.0
V u V p 0.2 is almost the same in the different width of plate. 0.0 0.5 1.0 1.5 2.0
Rp
The relation between V u V p 0.2 and R p considering the (b) Middle-joined plate
residual stress is shown in Fig. 4. The influence of the Fig. 4 Influence of residual stress
residual stress on the ultimate strength is small for both the
edge- and middle-joined plates. ıu/ıp0.2
1.0
ıu/ıp0.2
1.0
0.8
0.8
0.6
Non-joined plate
0.6
Non-joined plate 0.4 =200mm
0.4 b=200mm b
b =300mm =500mm
0.2 b
0.2 b =500mm Euler's buckling curve
Euler's buckling curve
0.0
0.0
0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0
Rp Rp
(a) Edge-joined plate ıu/ıp0.2
(a) Edge-joined plate
ıu/ıp0.2
1.0
1.0
0.8 0.8
0.6 0.6 b
Non-joined plate Non-joined plate
b
0.4 b=200mm 0.4 =200mm
b=500mm =500mm
0.2 0.2
Euler's buckling curve
Euler's buckling curve
0.0
0.0
0.0 0.5 1.0 1.5 2.0
Rp 0.0 0.5 1.0 1.5 2.0
Rp
(b) Middle-joined plate (b) Middle-joined plate
Fig. 3 Influence of joining locations Fig. 5 Influence of plate width
382
Transactions of JWRI, Vol.39 (2010), No. 2
References
[1] I. Okura, T. Nagao, T. Ishikawa, N. Hagisawa and S. Osumi: J.
of Structural and Earthquake Engineering, JSCE, 64 (2008),
pp.789-805.
383
Transactions of JWRI, Vol.39 (2010), No. 2
KEY WORDS: (Numerical analysis) (Nonlinear analysis) (Thermal Elastic Plastic analysis) (Dynamic Explicit
FEM) (Transient stress) (Transient deformation)
-200.0
of the limitation of the very short time increment. This
b
z B’
c 60.0 mm
y
limitation is due to the Courant condition, which is related 30 division
䞉䞉䞉
x 20.0 mm
to the stress wave propagation velocity and time increment. -400.0
10 division
The Courant condition requires the distance traveled by 0.0 10.0 20.0 30.0 40.0 50.0
stress wave in a time increment to be less than the minimum Time history (sec)
element size . In this study, the authors propose a method to
Fig. 1 Time history of stress ı x
overcome this limitation.
† Received on 30 September 2010 Transactions of JWRI is published by Joining and Welding
* Graduate School, Osaka Prefecture University, Osaka, Research Institute, Osaka University, Ibaraki, Osaka 567-0047,
Japan
Japan
384
Development of analytical method for welding mechanics using idealized explicit FEM
385
Transactions of JWRI, Vol.39 (2010), No. 2
(2) Using the model of bead-on-plate welding, the Idealized Explicit FEM and the static implicit FEM show almost
the same accuracy for transient stress, residual stress implicit FEM in the analysis of large-scale welding
and residual deformation. problem whose dof is 243243.
(3) By comparing the computing time, it is verified that the
Idealized Explicit FEM is 12 times faster than the static
386
CONTRIBUTIONS TO OTHER ORGANIZATIONS
(July 2010 ~ December 2010)
[Physics, Processes, Instruments & Measurements] T. IWAO, Y. MORI, M. OKUBO, T. SAKAI, S. TASHIRO,
M. TANAKA and M. YUMOTO
M. TANAKA, Y. TSUJIMURA, K. NAKATA, E. YAMAMOTO, Modelling of Metal Vapour in Pulsed TIG Including
K. YAMAZAKI and K. SUZUKI Influence of Self-Absorption
Dynamic Behavior of Metal Vapor in Arc Plasma J. Phys. D., Vol. 43 (2010), 434010.
during TIG Welding
63rd Annual Assembly of Int. Inst. Welding (IIW), S. TASHIRO, T. ZENIYA, K. YAMAMOTO, M. TANAKA,
(2010), IIW Doc. 212-1169-10. K. NAKATA, A. B. MURPHY, E. YAMAMOTO,
K. YAMAZAKI and K. SUZUKI
Y. TSUJIMURA and M. TANAKA Numerical Analysis of Fume Formation Mechanism in
A Numerical Model of GMA Welding from Arc Welding
Engineering View J. Phys. D., Vol. 43 (2010), 434012.
63rd Annual Assembly of Int. Inst. Welding (IIW),
(2010), IIW Doc. 212-1170-10. S. TASHIRO, T. ZENIYA, K. YAMAMOTO, M. TANAKA,
K. NAKATA, A. B. MURPHY, E. YAMAMOTO,
Y. TSUJIMURA, K. YAMAMOTO and M. TANAKA K. YAMAZAKI and K. SUZUKI
A Simulation Model of a Short Arc Lamp Numerical Analysis of Fume Formation Mechanism in
Ceram. Trans., Vol. 219 (2010), pp. 195-201. TIG Welding
Q. J. Jpn. Weld. Soc., Vol. 28, No. 4 (2010),
J. J. LOWKE, M. TANAKA and A. B. MURPHY pp. 369-375 (in Japanese).
Metal Vapour in MIG Arcs Can Cause (1) Minima in
Central Arc Temperatures and (2) Increased Arc S. TASHIRO, M. TANAKA and A. B. MURPHY
Voltages Numerical Analysis of Non-Equilibrium Plasma
Welding in the World, Vol. 54, No. 9/10 (2010), Property in Anode Boundary Layer of Argon Gas
pp. 292-297. Tungsten Arc
Surf. Coat. Technol., Vol. 205 (2010), pp. s115-s119.
M. TANAKA, K. YAMAMOTO, S. TASHIRO, K. NAKATA,
E. YAMAMOTO, K. YAMAZAKI, K. SUZUKI, A. B. MURPHY C. S. MOON, K. TAKEDA, M. SEKINE, Y. SETSUHARA,
and J. J. LOWKE M. SHIRATANI and M. HORI
Time-dependent Calculations of Molten Pool Surface Loss Probabilities of H and N Radicals on
Formation and Thermal Plasma with Metal Vapour in Different Materials in Afterglow Plasmas Employing
Gas Tungsten Arc Welding H2 and N2 Mixture Gases
J. Phys. D., Vol. 43 (2010), 434009. J. Appl. Phys., Vol. 107 (2010), pp. 103310-1
-103310-7.
A. B. MURPHY, M. TANAKA, K. YAMAMOTO, S. TASHIRO,
J. J. LOWKE and K. OSTRIKOV C. S. MOON, S. TAKASHIMA, K. TAKEDA, M. SEKINE,
Modelling of Arc Welding: The Importance of Y. SETSUHARA, M. SHIRATANI and M. HORI
Including the Arc Plasma in the Computational Etching Characteristics of Organic Low-K Films
Domain Interpreted by Internal Parameters Employing a
Vacuum, Vol. 85 (2010), pp. 579-584. Combinatorial Plasma Process in an Inductively
Coupled H-2/N-2 Plasma
M. MIYATA, S. TASHIRO and M. TANAKA J. Appl. Phys., Vol. 107 (2010), pp. 113310-1
Numerical Analysis of Plasma Keyhole Welding of an -113310-8.
Aluminum Thin Plate
Ceram. Trans., Vol. 219 (2010), pp. 189-194. Y. SETSUHARA, K. CHO, K. TAKENAKA, M. SHIRATANI,
M. SEKINE and M. HORI
S. TASHIRO, M. MIYATA, M. TANAKA, K. SHIN and Advanced Research and Development for Plasma
K. TAKAHASHI Processing of Polymers with Combinatorial
Numerical Analysis of Basic Property of Keyhole Plasma-Process Analyzer
Welding with Plasma Arc Thin Solid Films, Vol. 518 (2010), pp. 6320-6324.
Trans. Mater. Res. Soc. Jpn., Vol. 35, No. 3 (2010),
pp. 589-592.
387
Y. SETSUHARA, K. CHO, M. SHIRATANI, M. SEKINE and K. H. SONG and K. NAKATA
M. HORI Microstructural and Mechanical Properties of
X-Ray Photoelectron Spectroscopy Analysis of Friction-Stir-Welded and Post-Heat-Treated Inconel
Plasma–polymer Interactions for Development of 718 Alloy
Low-Damage Plasma Processing of Soft Materials J. Alloy. Compd., Vol. 505 (2010), pp. 144-150.
Thin Solid Films, Vol. 518 (2010), pp. 6492-6495.
J. LIAO, N. YAMAMOTO, H. LIU and K. NAKATA
Y. SETSUHARA, K. CHO, K. TAKENAKA, M. SHIRATANI, Microstructure at Friction Stir Lap Joint Interface of
M. SEKINE and M. HORI Pure Titanium and Steel
Low-damage Plasma Processing of Polymers for Mater. Lett., Vol. 64 (2010), pp. 2317-2320.
Development of Organic-Inorganic Flexible Devices
Surf. Coat. Technol., Vol. 205 (2010), pp. S355-S359. M. EGAWA, N. UEDA, K. NAKATA, M. TSUJIKAWA and
M. TANAKA
K. S. SHIN, Y. S. CHOI, I. S. CHOI, Y. SETSUHARA and Effect of Additive Alloying Element on Plasma
J. G. HAN Nitriding and Carburizing Behavior for Austenitic
Nano-crystalline Silicon Thin Films Grown by the Stainless Steels
Inductively Coupled Plasma Assisted CFUBM at Low Surf. Coat. Technol., Vol. 205 (2010), pp. s246-s251.
Temperature
Surf. Coat. Technol., Vol. 205 (2010), pp. S227-S230. M. AONUMA and K. NAKATA
Effect of Calcium on Intermetallic Compound Layer at
Y. SETSUHARA and M. HASHIDA Interface of Calcium Added Magnesium-Aluminium
Photon-induced Phonon Excitation Process as Alloy and Titanium Joint by Friction Stir Welding
Low-Temperature Nonequillibrium Nano-Surface Mater. Sci. Eng. B, Vol. 173 (2010), pp. 135-138.
Modification of Silicon
Surf. Coat. Technol., Vol. 205 (2010), pp. 1826-1829. H. LIU, K. NAKATA, N. YAMAMOTO and J. LIAO
Grain Orientation and Texture Evolution in Pure
K. TAKENAKA, Y. SETSUHARA, K. CHO, M. SHIRATANI, Titanium Lap Joint Produced by Friction Stir Welding
M. SEKINE and M. HORI Mater. Trans., Vol. 51, No. 11 (2010), pp. 2063-2068.
Combinatorial Analysis of Plasma-Surface Interactions
of Polyethyleneterephthalate with X-ray Photoelectron I. SEKI, H. KIMURA, K. NAKATA and A. INOUE
Spectroscopy Influence of Precipitation Behavior of Different
Jpn. J. Appl. Phys., Vol. 49 (2010), pp. 08JA02-1 Crystalline Phases for Embrittlement Behavior of
-08JA02-4. Several Zr-Based Metallic Glasses
Mater. Trans., Vol. 51, No. 11 (2010), pp. 2033-2038.
K. CHO, K. TAKENAKA, Y. SETSUHARA, M. SHIRATANI,
M. SEKINE, M. HORI, E. IKENAGA, H. KONDO, E. ASHIDA, X. ZHANG, S. TARASAWA and S. KATAYAMA
O. NAKATSUKA and S. ZAIMA Bead Formation and Stability in Low-Speed Welding
Hard X-Ray Photoelectron Spectroscopy Analysis for of Stainless Steel with High-Power Fiber Laser
Organic-Inorganic Hybrid Materials Formation J. Jpn. Laser Processing Soc., Vol. 16, No. 3 (2010),
Ceram. Trans., Vol. 219 (2010), pp. 183-188. pp. 165-170 (in Japanese).
388
M. MIZUTANI and S. KATAYAMA Y. KAWAHITO, Y. NIWA, T. TERAJIMA and S. KATAYAMA
Effect of Plume and Induced Hot Air above Specimen Laser Direct Joining of Glassy Metal Zr55Al10Ni5Cu30
on Laser Beam Refraction during Laser Welding to Engineering Plastic Polyethylene Terephthalate
IIW Commission IV, Vol. 2010, No. IV-1032-10 Mater. Trans., Vol. 51, No. 8 (2010), pp. 1433-1436.
(2010), pp. 1-11.
Y. ABE, M. MIZUTANI, Y. KAWAHITO and S. KATAYAMA
S. KATAYAMA Deep Penetration Welding with High Power Laser
Monitoring Apparatus and In-Process Control under Vacuum
Manufacturing by Laser, Practical Lecture Text for Proc. 29th Int. Congress on Applications of Lasers &
Chief Researchers and Engineers (2010), Electro-Optics (ICALEO 2010), (2010), pp.
pp. 5-65-5-89 (in Japanese). 648-653.
389
H. HONDA, M. TSUKAMOTO, N. ABE, T. SHINONAGA and A. KONDO, H. ABE, M. NAITO, Y. OKUNI, M. MIURA and
M. FUJITA N. ISU
Microstructures Formed on Stainless Steel by Novel Recycling Process of Waste FRP from
Femtosecond Laser Irradiation Advanced Materials
Ceram. Trans., Vol. 219 (2010), pp. 299-303. Ceram. Trans., Vol. 219 (2010), pp. 229-235.
T. NOMURA, M. TSUKAMOTO, N. ABE, M. TAKAHASHI and J. GUO, H. XU, S. WANG, S. ZHANG, H. ABE, X. WANG,
M. FUJITA Y. ZHAO and M. NAITO
Photoconductive Properties of Titanium Dioxide Film Eudragit Polymer Encapsulated Hydroxyapatite
by Femtosecond Laser Irradiation Microspheres as a Carrier for the Sustained Release of
Proc. 29th Int. Congress on Applications of Lasers & Small Molecular Drugs
Electro-Optics, (2010). Ceram. Trans., Vol. 219 (2010), pp. 265-272.
390
[Materials, Metallurgy & Weldability] S. LI, H. IMAI, H. ATSUMI and K. KONDOH
Phase Transformation and Precipitation Hardening
M. TAKAHASHI Behavior of Cr and Fe in BS40CrFeSn Alloy
Anodic Bonding - Its Principle and Potential for J. Mater. Sci., Vol. 45 (2010), pp. 5669-5675.
Application to Various Materials
NEW GLASS, Vol. 25, No. 3 (2010), pp. 7-12 S. LI, H. IMAI, H. ATSUMI and K. KONDOH
(in Japanese). Contribution of Ti Addition to Characteristics of
Extruded Cu40Zn Brass Alloy Prepared by Powder
M. WAHBA, S. KATAYAMA, T. KURODA and H. KIMURA Metallurgy
Laser Welding of Zn55Cu30Al10N5 Bulk Metallic Glass Mater. Des., Vol. 32 (2010), pp. 192-197.
Ceram. Trans., Vol. 219 (2010), pp. 73-79.
K. KONDOH, K. KANEKO and T. AKITA
S. KATAYAMA and M. MIZUTANI Advanced Powder Metallurgy Metals by Metal
Formation and Prevention of Porosity in Laser Working
Welding and Hybrid Welding Steel Res. Int., Vol. 81, No. 9 (2010), pp. 1283-1287.
Proc. 50 Years Memorial Symp. for Foundation of
Welding Metallurgy Research Committee of Japan H. IMAI, K. KONDOH, S. LI, H. ATSUMI, Y. KOSAKA and
Welding Society, -Advance and Innovation in study A. KOJIMA
of Welding Metallurgy-, (2010), pp. 24-27 Effect of Bismuth Addition on Machinability and
(in Japanese). Mechanical Properties of Lead-free Brass via Powder
Metallurgy Process
S. KATAYAMA Steel Res. Int., Vol. 81, No. 9 (2010), pp. 1296-1299.
Laser Joining of Metal and Plastic
Industrial Laser Solutions for Manufacturing, (2010), S. LI, H. IMAI, H. ATSUMI and K. KONDOH
pp. 15-16. Effects of Ti Addition on Microstructure and
Mechanical Properties of Extruded Cu40Zn-2.2Bi
S. KATAYAMA and T. OGAWA Brass by Powder Metallurgy
Laser Weldability and Aging Characteristics of Welds Steel Res. Int., Vol. 81, No. 9 (2010), pp. 1312-1315.
-Laser Weldability of Commercially Available A7N01
Alloy(1)- T. THRERUJIRAPAPONG, K. KONDOH, H. IMAI, J. UMEDA
J. Light Metal Welding and Construction, Vol. 48, and B. FUGETSU
No. 12 (2010), pp. 5-16 (in Japanese). Hot Extrusion of Pure Titanium Reinforced with
Carbon Nanotubes
M. WAHBA, M. MIZUTANI, Y. KAWAHITO and Steel Res. Int., Vol. 81, No. 9 (2010), pp. 1320-1323.
S. KATAYAMA
High-Brightness Disk Laser Welding of Magnesium S. LI, H. IMAI, H. ATSUMI and K. KONDOH
Alloys Precipitation Hardening Response of Ti Addition on
Proc. 29th Int. Congress on Applications of Lasers & BS40 Brass by Powder Metallurgy
Electro-Optics (ICALEO 2010), (2010), pp. CD Proc. World PM2010: Int. Powder Metallurgy
429-435. Congress & Exhibition, (2010), CD-ROM.
391
K. KONDOH, R. TAKEI, H. FUKUDA, J. UMEDA and H. IMAI T. MIYAZAWA, Y. IWAMOTO, T. MARUKO and H. FUJII
SKPFM Quantitative Evaluation of Initial Galvanic Development of Ir Based Tool for Friction Stir
Corrosion Phenomenon of Magnesium Alloys Welding of High Temperature Materials
Proc. Materials Science & Technology (MS&T) Q. J. Jpn. Weld. Soc., Vol. 28, No. 2 (2010),
2010, (2010), pp. 745-752. pp. 203-207 (in Japanese).
392
Y. TAKINAMI and S. KIRIHARA S. KIRIHARA, N. OHTA, T. NIKIL and M. KANEKO
Fabrication of Microwave Structures through Terahertz Wave Properties of Alumina Photonic
Dielectric Materials Percolation into Metal Patterns Crystals
J. Functionally Graded Mater., Vol. 20 (2010), Ceram. Trans., Vol. 216 (2010), pp. 121-127.
pp. 31-36 (in Japanese).
S. KIRIHARA, T. NIKIL and M. KANEKO
C. MAEDA, M. SUWA and S. KIRIHARA Terahertz Wave Harmonization in Geometrically
Fabrication of Artificial Bone Models with Graded Patterned Dielectric Ceramics Through Spatially
Porous Scaffold Structures Composed of Structural Joining
Hydroxyapatite Ceramics by Using Stereolithography Ceram. Trans., Vol. 216 (2010), pp. 113-119.
J. Functionally Graded Mater., Vol. 20 (2010),
pp. 37-40 (in Japanese). M. SUWA, S. KIRIHARA and T. SOHMURA
Freeform Fabrication and Structural Controls of
M. SHAHIEN, M. RADWAN, S. KIRIHARA, Y. MIYAMOTO Alumina Dental-Crown Models by Using
and T. SAKURAI Stereolithography
Combustion Synthesis of Single-Phase Ǻ-Sialons Ceram. Trans., Vol. 218 (2010), pp. 213-220.
(z=2-4)
J. Eur. Ceram. Soc., Vol. 30, No. 9 (2010), S. KIRIHARA, N. KOMORI and N. OHTA
pp. 1925-1930. Smart Processing for Ceramics Structure Tectonics:
Fabrication of Dielectric Micro Patterns for Artificial
N. OHTA, T. NIKI and S. KIRIHARA Photosynthesis in Terahertz Wave Regions by Using
Terahertz Wave Concentrations into Micro Reactors Stereolithography
Including Water Solvents Introduced into Alumina Adv. Sci. Technol., Vol. 63 (2010), pp. 141-146.
Photonic Crystals with Diamond Structures
J. Functionally Graded Mater., Vol. 20 (2010), Y. KOMIZO
pp. 13-18 (in Japanese). Welding Metallurgy I-2 Metallurgy for Welding of
Steel Materials
S. KIRIHARA, D. SANO, T. NIKI, N. OHTA and Textbook for summer school of welding engineering,
Y. TAKINAMI (2010), pp. 89-106 (in Japanese).
Fabrication of Micro Photonic Crystals for Handling of
Terahertz Electromagnetic Waves T. KANNENGIESSER, S. S. BABU, Y. KOMIZO and
Proc. Nanotech Conf. & Expo 2010, Vol. 2 (2010), A. J. RAMIREZ
pp. 223-226. In-situ Studies with Photons, Neutrons and Electrons
Scattering (Book)
S. KIRIHARA, N. KOMORI, M. NAKANO, N. OHTA and Published by Springer, (2010).
T. NIKI
Terahertz Wave Properties of Micro Patterned Titania D. ZHANG, H. TERASAKI and Y. KOMIZO
and Metallic Glass Particles in Hexagonal Tablet In Situ Observation of Phase Transformations in
Fabricated Using Microstereolithography Fe-0.08C Alloy
Ceram. Trans., Vol. 219 (2010), pp. 61-66. Ceram. Trans., Vol. 219 (2010), pp. 397-402.
393
H. TERASAKI T. TAKAHASHI, S. KOMATSU, H. NISHIKAWA and
In-situ Studies with Photons, Neutrons and Electrons T. TAKEMOTO
Scattering (Book) High-Temperature Resistant Intermetallic Compound
Published by Springer, (2010). Joints for Si Chips and Cu Substrates
J. Electron. Mater., Vol. 39, No. 10 (2010),
T. DAITO, H. NISHIKAWA, T. TAKEMOTO and pp. 2274-2280.
T. MATSUNAMI
Impact Test of Sn-3.0Ag-0.5Cu (-xCo) Solder with W. CHEN, Y. MIYAMOTO, T. MATSUMOTO and T. TOJO
Co-P Plating Spark Plasma Sintering of Carbon Particles Coated
Proc. Int. Conf. on Electronics Packaging 2010 with Al2O3 Ceramic
(ICEP 2010), (2010), pp. 796-799. Ceram. Trans., Vol. 219 (2010), pp. 285-291.
394
A. KOBAYASHI, T. KURODA, H. KIMURA and A. INOUE K. TOKUNAGA, N. TSUKUDA, N. YOSHIDA, Y. YAHIRO,
Microstructure and Properties of a Ni-Based Metallic M. MITSUHARA, H. NAKASHIMA, A. KOBAYASHI,
Glass Coating Produced by Gas Tunnel Type Plasma K. EZATO, S. SUZUKI and M. AKIBA
Spraying Tungsten Coatings on Reduced-Activation
Ceram. Trans., Vol. 219 (2010), pp. 37-42. Ferritic/Martensitic Steel by Plasma Spray Technique
Appl. Plasma Sci., Vol. 18, No. 2 (2010), in press
A. KOBAYASHI, T. KURODA, H. KIMURA and A. INOUE (in Japanese).
Microstructure and Properties of Zr-Based Metallic
Glass Coatings by Gas Tunnel Type Plasma Spraying
Appl. Plasma Sci., Vol. 18, No. 1 (2010), pp. 43-50 [Mechanics, Strength & Structural Design]
(in Japanese).
A. MAEKAWA, K. NAKACHO, N. MA and R. SATO
T. KURODA, A. KOBAYASHI and K. IKEUCHI Residual Stresses Measurement of Large-Bore
Microstructure and Toughness for Resistance Butt Stainless Steel Pipe with Butt-Welded Joint by
Welding of Super Duplex Stainless Steel Inherent Strain Method
Appl. Plasma Sci., Vol. 18, No. 1 (2010), pp. 57-63 Proc. 2010 Annual Meeting of JSME/MMD, (2010),
(in Japanese). pp. 1260-1262 (in Japanese).
395
Y. AGANO, S. FUJIHIRA, S.-H. LEE and Y.-C. KIM Y.-C. KIM, M. HIROHATA and K. INOSE
Experimental Study on Repair Welding under Static Effects of Phase Transformation on Distortion and
and Cyclic Loads Residual Stress Generated by LBW on High Strength
Proc. Int. Conf. on Bridge Maintenance, Safety and Steel
Management, Life-Cycle Optimization, CD-ROM 63rd Annual Assembly of Int. Inst. Welding (IIW),
(2010), pp. 2998-3004. (2010), IIW Doc. XV-1353-10, IIW Doc.
IV-1033-10.
M. HIROHATA and Y.-C. KIM
Identification of Factor Determining Strength under Y. SAKINO, T. KURODA, A. KOBAYASHI and Y.-C. KIM
Compressive Loads of Steel Members Corrected by Fracture Behavior of Zr48Cu36Al8Ag8 and
Heating/Pressing Zr55Cu30Al10Ni5 Bulk Metallic Glasses in High Speed
Int. J. Steel Structures, Vol. 10, No. 3 (2010), Tensile Test
pp. 245-251. Front. Appl. Plasma Technol., Vol. 3, No. 2 (2010),
pp. 102-106.
J.-Y. LEE, K. INOSE and Y.-C. KIM
Verification of Validity and Generality of Dominant Y. SAKINO, T. KURODA and Y.-C. KIM
Factors in High Accuracy Prediction of Welding Tensile Mechanical Behavior of Zr-based Bulk
Distortion Metallic Glass in High-speed Tensile Tests
Welding in the World, Vol. 54, No. 9/10 (2010), Ceram. Trans., Vol. 219 (2010), pp. 43-47.
pp. R279-R285.
Y. SAKINO and Y.-C. KIM
D.-H. PARK, M. HIROHATA and Y.-C. KIM Effect of Welded Condition on Charpy Absorbed
Effects of 3-D Welding Residual Stress on Energy of Heat Affected Zones in Low Toughness
Elastic-plastic Behavior of Steel Structural Member Steel
J. Constructional Steel, Vol. 18 (2010), pp. 151-158 Steel Constr. Eng., Vol. 17, No. 67 (2010),
(in Japanese). pp. 43-52.
Y. AGANO, S.-H. LEE, S. FUJIHIRA and Y.-C. KIM Y. SAKINO, S. TAKAHASHI and Y.-C. KIM
Effects of Static Loads on Weldability Effects of Strain Rate on Tensile Strength of Steel
—Repair/Reinforcement by Welding for Steel Bridges Specimens of HAZs with Stress Concentrations
in Service— Q. J. Jpn. Weld. Soc., Vol. 28, No. 3 (2010),
Q. J. Jpn. Weld. Soc., Vol. 28, No. 4 (2010), pp. 328-337 (in Japanese).
pp. 391-394 (in Japanese).
Y. SAKINO, K. YOSHIKAWA, Y. SANO and Y.-C. KIM
Y. AGANO, S.-H. LEE, S. FUJIHIRA and Y.-C. KIM Residual Stress of High Strength Steel and Its Welds
Verification of Weldability under Cyclic Loads Generated by Laser Peening
—Repair/Reinforcement by Welding for Steel Bridges J. Constructional Steel, Vol. 18 (2010), pp. 493-498
in Service— (in Japanese).
Q. J. Jpn. Weld. Soc., Vol. 28, No. 4 (2010),
pp. 436-442 (in Japanese). Y. SAKINO and Y.-C. KIM
Effect of Multilayered Thermal Cycle on Charpy
D.-H. PARK, M. HIROHATA, S. TAMAGAWA and Y.-C. KIM Absorbed Energy in Heart Affected Zone
Introducing/Reproducing of Welding Distortion and J. Constructional Steel, Vol. 18 (2010), pp. 253-258
Residual Stress in Analysis for Elastic-Plastic (in Japanese).
Behaviors of Steel Plates under Compressive Loads
Int. J. Steel Structures, Vol. 10, No. 4 (2010), T. TERAJIMA, F. TAKEUCHI, K. NAKATA, S. ADACHI,
pp. 349-357. K. NAKASHIMA and T. IGARASHI
Composite Coating Containing WC/12Co Cermet and
D.-C. PARK, M. HIROHATA and Y.-C KIM Fe-based Metallic Glass Deposited by High-Velocity
Consideration for Welding Imperfection in Oxygen Fuel Spraying
Elastic-plastic Large Deformation Analysis J. Alloy. Compd., Vol. 504 (2010), pp. S288-S291.
63rd Annual Assembly of Int. Inst. Welding (IIW),
(2010), IIW Doc. XV-1352-10.
396
[General Welding]
S. KATAYAMA
Laser Welding of Dissimilar Materials
Rev. Laser Eng., Vol. 38, No. 8 (2010), pp. 594-602
(in Japanese).
397
Editorial Members
Chairman ···································· NAITO Makio
TANAKA Manabu
KATAYAMA Seiji
KIM You-Chul
FUJII Hidetoshi
ABE Hiroya
Published by
Joining and Welding Research Institute,
Osaka University
11-1 Mihogaoka, Ibaraki, Osaka 567-0047, Japan
Telephone (06) 6877-5111
Facsimile (06) 6879-8689
Printed by
SEIEI Printing Co., Ltd.
2-1-33, GAMO, Joto-ku, OSAKA