No. 30 DEC.

2011
Feature-Ⅰ: Material Processing Technologies
Feature-Ⅱ: Steel Wire Rod and Bar
Feature-Ⅲ: Steel Plate and Sheet

Contents
Feature-Ⅰ Material Processing Technologies
High-quality Work Roll Manufacturing Technology Using New Electro
Slag Remelting (ESR)…………………………………………………………………… 1
Jun SATO, Kouji IWANAGA, Atsushi TOMIOKA, Katsushige NISHIGUCHI, Hiroki NAKASHIMA, Hitoshi ISHIDA
Predicting Effect of Cold Rolling on Fatigue Strength under Combined Loading … 7
Mariko MATSUDA, Eiji OOTSUKI, Shuhei KAJIHARA, Yoji HANAWA, Takeshi HAMADA
Effect of Alloying Elements on Machinability and Hot Workability of α-β
Titanium Alloy Containing Fe and C ………………………………………………… 13
Shogo MURAKAMI, Dr. Katsuhiko OZAKI, Kousuke ONO, Yoshio ITSUMI
Pre-coated Titanium Sheet with Excellent Press Formability …………………… 19
Akihisa FUJITA, Yoshio ITSUMI, Tadashige NAKAMOTO, Kayo YAMAMOTO

Editor-in-chief : Newly Developed Iron-based Powder Mixture, High-density SEGLESS
,

Yuichi SEKI for High Density Compaction ………………………………………………………… 24
Hironori SUZUKI, Satoshi NISHIDA, Takayasu FUJIURA

Associate Editors : Newly Developed Iron Powder for Highly Efficient Dust Cores ………………… 30
Hirofumi HOJO, Nobuaki AKAGI, Tetsuya SAWAYAMA, Hiroyuki MITANI
Tomokazu NAKAGAWA
Hidetoshi INOUE Feature-Ⅱ Steel Wire Rod and Bar
Overseas Deployment of Wire Rod & Bar Manufacturing and Wire Rod
Editorial Committee :
Secondary Processing ………………………………………………………………… 36
Toru HASHIMURA Goro AKAISHI, Noriaki HIRAGA
Kenichi INOUE
Development History of Wire Rods for Valve Springs …………………………… 41
Yasushi MAEDA Nao YOSHIHARA
Tsuyoshi MIMURA Development Trends of Soft Magnetic Iron ………………………………………… 46
Tsuneaki NISHIKAWA Dr. Masamichi CHIBA
Hiroyuki SHIMIZU Influence of Ti Precipitate in Carburizing Steel Containing Boron ……………… 52
Noboru TAKENOSHITA Dr. Nariaki OKAMOTO, Yosuke SHINDO, Mutsuhisa NAGAHAMA
Syoji YOSHIMURA Low Alloy Steel for Fracture Splitting Connecting Rod …………………………… 57
Akihiro MATSUGASAKO
Production Technology of Wire Rod for High Tensile Strength Steel Cord ………62
Kazuhiko KIRIHARA

Feature-Ⅲ Steel Plate and Sheet

Characteristics of Brittle Crack Arrest Steel Plate for Large Heat-input
Published by Welding for Large Container Ships ………………………………………………… 66
Masahito KANEKO, Dr. Tokutaka TANI
Secretariat & Publicity Dept. Effect of Steel Toughness on Brittle Crack Arrest Behavior of T-weld Joint
Kobe Steel, Ltd.
Structure Using Thick Plates ………………………………………………………… 70
Dr. Eiichi TAMURA, Dr. Tomokazu NAKAGAWA, Kazuyuki TSUTSUMI, Naohiro FURUKAWA
10-26, Wakinohamacho 2-chome, Chuo-
ku, Kobe, HYOGO 651-8585, JAPAN Effects of Ca Addition on Formation Behavior of TiN Particles and HAZ
http://www.kobelco.co.jp Toughness in Large-Heat-Input Welding …………………………………………… 76
Taku KATO, Shinsuke SATO, Dr. Hiroki OHTA, Toyoaki SHIWAKU
Editorial Office: Shinko Research Characteristics of Highly Formable 590-980MPa Grade Hot-dip Galvannealed
Co., Ltd.
5-5, Takatsukadai 1-chome,
Steel Sheets for Automobiles ………………………………………………………… 80
Dr. Yuichi FUTAMURA, Masaaki MIURA, Michitaka TSUNEZAWA
Nishi-ku, Kobe, HYOGO 651-2271, JAPAN
Fax: +81-78-992-9790 Formability of TRIP Type Banitic Ferrite Steel Sheet ……………………………… 85
E-mail: rd-office@kobelco.com Takayuki KIMURA
Integrated Application Method for KOBEHONETSUTM Steel Sheet ……………… 90
c Kobe Steel, Ltd. 2011
⃝ Dr. Yasuo HIRANO, Tetsuya IGARASHI, Haruyuki MATSUDA, Makoto NISHIMURA
 High-quality Work Roll Manufacturing Technology
Using New Electro Slag Remelting (ESR)
Jun SATO＊1, Kouji IWANAGA＊1, Atsushi TOMIOKA＊1, Katsushige NISHIGUCHI＊1, Hiroki NAKASHIMA＊2, Hitoshi ISHIDA＊3
＊1
Steel Making & Casting Department, Steel Casting & Forging Plant, Steel Casting & Forging Division, Iron & Steel Business, ＊2 Technical
Development Department, Steel Casting & Forging Division, Iron & Steel Business, ＊3 Materials Research Laboratory, Technical Development Group

Because cold rolling requires work rolls of high quality Bottom L Top
r
in their surfaces and interiors, the rolls are generally
made from electro-slag-remelting (ESR) ingots which
ensure a stable outcome. In order to produce rolls with θ1 θ2

excellent dendrite pattern, a fine and uniform dendrite Use range in roll
structure with no flow pattern is required. Segregation Freckle
θ Dendrite angle
lines, which may appear as "freckles" on the roll surface, good no good
are sometimes generated in the ingots. Such segregation Surface pattern
lines should not exist within the use depth of the rolls. Fig. 1 Necessary quality for roll material
In recent years, Kobe Steel has upgraded its ESR
apparatus and achieved a significant improvement in the slag layer and sequentially solidifies into a steel
the quality of the surface and interior. ingot. The steel ingot thus obtained has a superior
cleanliness and is used for high-value-added
Introduction products such as work rolls and aircraft members.
The quality of work rolls is largely affected by the
Work rolls are used for cold-rolling steel plates quality of the steel ingots made by the ESR process.
into thin sheets and have to meet demanding Here, the dendrite structure of the ingot surface must
standards for their surface and interior quality. Thus, be fine and homogeneous, and, to prevent freckles,
they are typically made from steel ingots produced there should be no positive segregation of the
by electro slag re-melting (ESR), ingots which can additives in the steel within the use range (depth).
consistently provide the desired quality. To ensure Depending on the intended applications, the rolls
high resistance to surface deterioration, the surfaces have differing quality requirements. Tinplate roll
must have fine and homogeneous dendrite for beverage cans especially requires fine and
structures without flow patterns. Some ingots may homogeneous dendrite structures on the rolls' surfaces.
have internal segregation lines, which can appear as In addition to the stringent quality requirement for
"freckles" on their surfaces, depending on the type of the surface, some such rolls are reworked for
steel. Such segregation lines must not exist in the reclamation after use and must be free from freckles
effective use depth of the rolls. To meet such even after twice as deep a layer is removed from the
demanding quality requirements，Kobe Steel renewed surface, compared with other types of rolls.
its ESR apparatus in 2007 and has significantly In order to refine the structure of surface dendrite,
improved the surface and interior qualities, as a large enough angle must be retained between the
reported in this paper. mould wall and the dendrite growth direction in a
steel ingot (the angle hereinafter referred to as
1．Method for making work rolls for cold rolling "dendrite angle"). This dendrite angle enables the
and characteristics required for rolls evaluation of the surface structure of rolls. The
present paper uses dendrite angle θ as the evaluation
The work rolls for cold rolling (hereinafter simply index of the surface quality and depth of the freckled
referred to as "work rolls") are made according to the region (hereinafter referred to as freckled depth, L / r
following steps: (L; the distance between the ingot circumference and
Melting by electric furnace (EF) → Refining in the depth where freckles appear: r; steel ingot radius)
ladle furnace (LF) → Bottom teeming & Ingot making as the evaluation indices of interior quality (Fig. 1). The
→ ESR → Forging & Heat treatment → Machining following describes the improvement of these indices.
ESR is a type of re-melting technique which involves
a target metal used as a consumable electrode, and a 2. Relationship between dendrite angle and freckled
water-cooled mould holding a layer of molten slag, depth
in which the slag layer conducts a large electric
current and generates resistance heat to melt the Dendrites grow vertically from the solidification
electrode. The droplet of molten steel passes through interface in the pool of molten steel 1), 2). Thus, in order

1 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

to increase the dendrite angle, the lateral face of the the melting rate alone cannot improve the dendrite
molten steel pool must be kept parallel to the angle and freckled depth at a time.
mould surface as much as possible. In general, the
higher the melting rate, the deeper the molten steel 3. Decreasing thickness of slag skin
pool becomes, which increases the dendrite angle.
Freckles, on the other hand, are caused by condensed In addition to the melting rate, various other
molten steel floating up and being trapped in the melting conditions can affect the quality of steel
solidified layer. Thus, the shallower the molten ingots made by ESR; e.g., flux composition, the
steel pool, the less likely they are to appear 3), 4). To amount of flux, electric current, voltage, cooling
maintain the pool of molten steel shallow enough, capacity of the mould and fill ratio. Aiming at
the melting rate must be kept low. In other words, improving both the dendrite angle and freckled
there is a trade-off relationship between the dendrite depth simultaneously, Kobe Steel focused on the heat
angle and freckled depth: i.e., a higher melting rate transfer of the slag skin formed between the ingot
results in a larger dendrite angle θ and shallower and water-cooled copper mould. A method was
L/r, and vice versa (Fig. 2). Fig. 3 shows the devised for thinning the slag skin to increase the
relationship between the dendrite angle θ and cooling rate of the steel ingot and for making the
freckled depth (L/r). Here, the dendrite angle is lateral face of the pool of molten steel almost parallel
defined by the average of the angles measured at to the mould surface to decrease the overall depth of
the depth of 30mm from the steel ingot surface the pool (Fig. 4).
(assuming that the measured points will lie on the The thickness of the slag skin is largely affected by
outer surface at the time of shipment) and the angle the slag composition and immersion volume of the
measured at the depth of 90mm (assuming that the electrode. Thus, for thinning the slag skin, it is
points will lie on the outer surface at the time of effective to use flux with a low melting point and/or
disposal). This figure also indicates that controlling to decrease the immersion volume of the electrode.
Such a flux can be found in a CaO-Al2O3-CaF2 series
with high CaF2 content. However, when used in an
insulated system in which the mould and steel ingot
Electrode
are insulated from each other, the flux can result in
arcing caused by the potential difference between the
r mould and steel ingot. Thus such flux cannot be used
Molten slag L
θ for the conventional apparatus. To enable the use of
θ
Molten flux with such a low melting-point, the mould and
Water cooled metal
copper
steel ingot must have the same electrical potential
Ingot (Fig. 5). The new ESR adopts a live mould which
allows electrical conduction and enables the use of
Range of freckle generation flux with a low melting point (1,290℃).
Melting rate High Low Immersing a large volume of electrode into the
Molten metal depth Deep Shallow
Dendrite angle (θ) Big(good) Small(no good)
slag layer increases the amount of heat extracted by
Depth of freckled region (L/r) Small(no good) Big(good) the electrode, which lowers the temperature of the
Fig. 2 Relationship between melting rate, molten metal depth, slag layer and increases the thickness of the slag skin.
dendrite angle (θ), and depth of freckled region (L /r ) Thus it is preferable to keep the immersion volume of

0.5 Surface of metal pool

Depth of freckled region（L/r）

Superior before improving

0.4
Electrode
Low melting rate
0.3

0.2 Molten slag

High melting rate
θ
0.1 Molten Shallow
metal
0
0 10 20 30 40 50
Dendrite angle（θ） Ingot
Fig. 3 Relationship between dedrite angle and depth of
freckled region（L /r ） Fig. 4 Ideal shape of molten metal

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 2

 Insulated mould
Table 1 Electrode positoning control
Live mould
Mould Mould Conventional ESR New ESR
Electrode Electrode
Electric Electric
Electrode positioning control Voltage swing Resistance swing
current current
Molten slag Molten slag Electrode positioned by Winch hoisting Ball screw

Molten metal Molten metal Amount of immersion 10mm 1∼3mm

Spark

Melting point of flux Melting point of flux 4. Quality of steel ingots produced by new ESR
：1,330℃ ：1,290℃

Fig. 5 Pattern diagrams of insulated mould and live mould in The following evaluates the quality of steel ingots
ESR melting
produced by conventional and new ESR apparatuses
and discusses the optimum operating conditions.
the electrode as small as possible. This was difficult
to achieve in the conventional apparatus, in which 4.1 Experimental method
the electrode was suspended by a wire with poor
responsivity, making it difficult to precisely control A steel ingot having an extra portion 500mm in
the electrode position. In addition, the conventional length at the top was prepared by the new ESR.
electrode positioning system, involving voltage swing, Specimens, as shown in Fig. 7, were prepared for
suffered from difficulties caused by slag composition, macroscopic observation: i) a φ810×30mm transverse
electrical conductivity and slag volume, all of which cross-section specimen; and ii) a 810×470×30mm
change during the process. In addition, the changing longitudinal section specimen.
lengths of electrode and ingot accompany changes in Among the segregation spots observed on the
the impedance and voltage, causing difficulties. To transverse cross-section specimen, spots having
avoid this, a resistance swing method was employed average diameters greater than 2mm were determined
for the new electrode positioning (Table 1). Fig. 6 to be freckles, and their locations were recorded.
summarizes these and other features of the ESR The angles of dendrite growth directions were
apparatus before and after the renewal. determined on the longitudinal section.

Conventional Motor New

Motor System with high-precision ball screw

Winch hoisting of electrode Revision of electrode positioning control system

 Voltage swing control ・Adoption of resistance swing control
・Revision of drive system

Electric current
Electric current

Coaxial feed system

(axisymmetric electromagnetic field)

Noncoaxial feed system Enhanced mould cooling

High slag temperature

・slag with high heat conductivity and
 low viscosity (high electric conduction)

Adoption of live mould

 (providing conductivity:
Feed back of electric current
  electric supply to mould)
 due to the insulation system

Insulated Conductive

Fig. 6 Features of conventional and new ESR

3 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

 500mm 0.5

Superior

Depth of freckled region (L/r)

810mm Extra material Product 0.4
for investigation

TOP BOTTOM 0.3

500mm

0.2

30mm
① 810mm 0.1
② Conventional ESR
New ESR
0
0 10 20 30 40 50
30mm
Dendrite angle (θ)
Fig. 7 Specimens for investigation of ESR
Fig.10 Relationship between dendrite angle (θ) and depth of
freckled region (L /r ) in the ingot produced by New
ESR

5. Theoretical verification of improvement in

internal quality of steel ingots

The present modification of the ESR apparatus

has changed the conventional casting conditions,
such as control method, slag composition, slag
r＝405mm temperature, molten steel temperature, slag skin
Freckle thickness and mould structure. Thus it is envisaged
L＝150mm that the solidification profile has also been changed.
To theoretically verify the quality improvement of
Fig. 8 Photograph of typical cross-sectional macro structure
the steel ingot achieved thanks to this modification,
〔Freckled region:150mm deep from the surface a numerical analysis (heat-transfer calculation) was
(L /r ＝0.370)〕 conducted to determine the shape of the molten steel
pool (i.e., dendrite angle). The calculation result was
compared to the results for the steel ingot quality
described in section 4.
Dendrite angle
An FEM program for solidification analysis,
"Casting Analysis System" (CASTEM), developed
θ by Kobe Steel5), was used for calculating the
Dendrite growth direction advancement of the solidification interface and the
shape of the molten steel pool in the ESR. The
CASTEM allows resolving thermal conduction
equations in non-steady state, taking into account the
Fig. 9 Photograph of typical longitudinal sectional macro
structure release of latent heat during solidification, as follows:
∂T ∂fs ∂ ∂T ∂ ∂T ∂ ∂T
ρC −ρL ＝ λ ＋ λ ＋ λ
   ∂t ∂t ∂x ∂x ∂y ∂y ∂z ∂z
4.2 Results wherein T represents temperature, t represents time,
C represents specific heat, ρ represents density, L
Fig. 8 and 9 show typical macroscopic structures represents latent heat, fs represents solid phase ratio
observed on transverse and longitudinal cross- and λ represents thermal conductivity.
sections, respectively. Fig.10 shows the relationships Table 2 shows the physical properties used for
between the dendrite angle and freckled depth in the heat-transfer analysis. The melting rates were
the ingots prepared by the new and conventional set to three conditions, i.e., 600kg/h, 700kg/h and
ESR apparatuses, respectively. It should be noted that 920kg/h. Since there is no measurement data for the
the new system has improved both the dendrite angle melt temperature T (℃) at the surface of the molten
and freckle depth significantly compared with the steel pool, the melt temperature was calculated
conventional system. according to Formula (1). A slag temperature measured
for a different type of steel produced by the

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 4

 Table 2 Physical properties and experimental conditions （×10−4）
used in heat transfer analysis 10.0
New ESR slag Conventional slag
Liquidus
1,450℃
temperature Chemical composition of work roll materials
8.0

Overall heat transfer coefficient

Solidus for cold rolling
1,344℃ (wt％)
temperature
C Si Cr Mo

(cal/(mm2・s・℃))
Liquid steel Heat −2 2 0.9 0.6 5.0 0.3
3.8×10 cal/
（mm・s・℃） 6.0
physical conductivity
properties Slag composition
(wt％)
Specific heat 0.17cal/g/K CaF2 CaO Al2O3 SiO2
−3 3
Density 7.6×10 g/mm Conventional ESR 25 34 34 7
4.0 New ESR 40 30 30 −
Latent heat of
65cal/g
solidification
−3 2 Conventional ESR
Ingot bottom 1.4×10 cal/
（mm・s・℃） 2.0
Coefficient of
heat transfer Ingot side −4 2
2.7×10 cal/
（mm・s・℃）
surface New ESR
Melting speed 600kg/h 700kg/h 920kg/h 0.0
0 2 4 6 8 10
Molten metal temperature 1,604℃＋α 1,628℃＋α 1,681℃＋α Slag skin thickness ds (mm)
Fig.11 Slag skin thickness and overall heat transfer coefficient
conventional ESR was used for the calculation. Here,
α represents a fitting parameter. Table 3 Slag skin thickness and overall heat transfer
T (℃)＝0.24×(melting rate kg/h) +1460 +α… (1) coefficient
The overall heat transfer coefficient h between Overall heat transfer
Slag skin thickness
coefficient
the lateral face of the steel ingot and the water-cooled (mm) −4 2
（×10 cal/ （mm ・s・℃））
copper mould was calculated using Formula (2) Conventional ESR 3 1.05
below. New ESR 1 2.68
ingot slag Cu Cooling
skin mould water
h1 h2 h3
λs λc Liquidus line Solidus line
Dendrite angle
ds dc 2,000 2,000 2,000
θ 753min. 499min.
h 618min.
1,800 710min. 1,800 1,800 463min.
1 1 1 1 1 1 430min.
＝ ＋ ＋ ＋ ＋ ……………… (2) 1,600 620min 1,600 551min. 1,600
   h h1 λs/ds h2 λc/dc h3 Depth 30mm
1,400 549min. 1,400 1,400 Depth 90mm
Fig.11 shows the relationship between the slag 430min.
Height (mm)

Height (mm)

Height (mm)
1,200 1,200 1,200 297min.
skin thickness and overall heat transfer coefficient 431min.
1,000 1,000 1,000
calculated under the above conditions. It appears 292min.
800 800 800
that the slag skin thickness for the conventional 294min.
600 600 600 156min.
ESR is approximately 3mm, while that for the new
155min.
400 157min. 400 400
ESR is approximately 1mm, and the overall heat 84min.
200 83min. 200 84min. 200
transfer coefficient is considered to have increased by
0 0 0
approximately 2.5 times (Table 3). 0 200 400 0 200 400 0 200 400
Radius(mm) Radius(mm) Radius(mm)
Fig.12 shows examples of the metal pool shapes (a) 600kg/h (b) 700kg/h (c) 920kg/h

calculated. The calculation is based on a two Fig.12 Metal pool shapes at each melting speed (α＝50℃)
dimensional axial symmetry model with the left
side representing the ingot center and the right side
representing the ingot surface. Each diagram Measured value Calculated value (30mm from surface)
Calculated value (90mm from surface) Calculated average
includes metal pool shapes, which are determined
45
by the temperature profiles of liquidus line and
40
Average dendrite angle (°)

solidus line at given times.

Dendrite growth directions were calculated at the 35

depth positions of 30mm and 90mm from the ingot 30

surface, in the direction vertical to the molten metal 25
pool determined by the profiles of the liquidus line. 20
Dendrite angles are defined by the crossing angles
15
between the dendrite growth directions and the ingot
10
surface. The dendrite angles were determined by the 500 600 700 800 900 1,000
average of angles obtained from the depth positions Melting speed (kg/h)
of 30mm and 90mm. Fig.13 Calculated (Molten metal temperature α＝0℃) and
Fig.13 compares the calculated and measured measured values of average dendrite angle

5 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

values of average dendrite angles. The figure indicates The above discussion indicates that the new ESR
the difference between the calculated and measured apparatus maintains the slag temperature 50~100℃
values. It should be noted that the values in Fig.13 higher than the conventional apparatus. This
were calculated using the slag temperature measured produces a slag skin 1mm thick, which is less than
for the conventional apparatus. The new apparatus the slag skin thickness of 3mm produced by the
is thought to have achieved a higher slag layer conventional apparatus. The thinner slag skin
temperature (= molten metal temperature) as a result increases the cooling rate at the lateral face of the
of the present renewal including the modification of ingot, which improves the shape of the molten steel
electrode positioning control (employing a resistance pool (Figure 4) and produces a steel ingot with a
swing control and a new electrode drive method), large dendrite angle and fewer freckle defects.
which decreased the immersion volume of the
electrode. 6. Summary
Thus, the fitting parameter α in Formula (1) was
set to 0℃, 50℃ and 100℃ to determine the molten An ESR apparatus was renewed so as to produce
metal temperature T (℃) at the pool surface for each high-quality rolls. The new apparatus was subjected
condition. The dendrite angles were averaged for to a study of the relationship between the melting
each condition and were compared to measured conditions and the quality of the steel ingot interior.
values as shown in Fig.14 and 15 respectively. The following results were obtained:
The comparison with measured values indicates (1) Significant improvements were obtained for the
that, by setting α to a value in the range from dendrite angle and depth of the freckled region in
50~100 ℃, the calculated values for dendrite angles the new ESR, compared with the conventional
turn out to agree well with the measured values. one;
(2) the new ESR yielded a thinner slag skin of 1mm,
as against the 3mm thick slag skin in the
Measured value Calculated value (30mm from surface)
Calculated value (90mm from surface) Calculated average conventional ESR, and increased the overall
45 coefficient of heat conduction by approximately
40
2.5 times; and
Average dendrite angle (°)

(3) the results of the studies on the quality of the

35
ingot interior and solidification analysis indicate
30
that the new ESR increased the slag temperature
25 by 50~100 ℃, compared with the conventional
20 ESR. It is contemplated that the higher slag
15 temperature has enabled the thinning of the slag
10 skin.
500 600 700 800 900 1,000
Melting speed (kg/h)
Conclusions
Fig.14 Calculated (Molten metal temperature α＝50℃) and
measured values of average dendrite angle
The quality evaluation and its results on the roll
material produced by a new ESR apparatus have
Measured value Calculated value (30mm from surface)
Calculated value (90mm from surface) Calculated average
been outlined. The new apparatus further comprises
45 an automatic melt control system which decreases
40
the operator dependence of the outcome, enabling
Average dendrite angle ( °)

consistent production of high quality roll materials.

35

30
References
25

20 1) A. Suzuki et al., Physical Chemistry of Electro Slag Remelting

and Its Technical Issues (1979), pp.132-133.
15
2) A. Masui et al., Physical Chemistry of Electro Slag Remelting
10 and Its Technical Issues, pp.149-151.
500 600 700 800 900 1,000
3) H. Yamada et al., Tetsu-to-Hagane，No. 1 (1989), pp.97-104.
Melting speed (kg/h)
4) H. Yamada et al., Tetsu-to-Hagane，No. 1 (1989), pp.105-112.
Fig.15 Calculated (Molten metal temperature α＝100℃) 5) T. So et al., R&D KOBE STEEL ENGINEERING REPORTS,
and measured values of average dendrite angle Vol. 37, No. 4 (1987), pp.99-100.

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 6

 Predicting Effect of Cold Rolling on Fatigue Strength
under Combined Loading
Mariko MATSUDA＊1, Eiji OOTSUKI＊2, Shuhei KAJIHARA＊3, Yoji HANAWA＊4, Takeshi HAMADA＊5
＊1
Technical Development Department, Steel Casting & Forging Division, Iron & Steel Business
＊2
Forging Department, Steel Casting & Forging Plant, Steel Casting & Forging Division, Iron & Steel Business
＊3
Mitsui Engineering & Shipbuilding Co., Ltd.
＊4
Mechanical Engineering Research Laboratory, Technical Development Group
＊5
Mechanical Engineering Research Laboratory, Technical Development Group (currently with KOBELCO CRANES CO., LTD.)

The cold rolling method is adopted to improve the predictable the improvement in fatigue strength that
fatigue strength of crankpin fillet for marine diesel is an effect of the compressive residual stress applied
engines. Under working conditions, combined bending by cold rolling. Fatigue tests were conducted under
and torsional loading occurs in crankpin fillet; combined load conditions with mean stress. Three
therefore, combined axial and torsional fatigue tests methodologies were selected from among the
under tensile or compressive mean stress were conducted various conventional approaches that have been
to distinguish among the three methods of fatigue proposed for evaluating fatigue strength under
strength evaluation. Modified IACS and Findley combined loading. The three methodologies were
criteria were confirmed to be methods that are on the evaluated for their validity. Furthermore, a new
safe side. By evaluating fatigue strength by these criteria, method for evaluating fatigue strength under
it was confirmed that compressive residual stresses combined load was adopted for the pin fillet of a
from the cold rolling method improve fatigue strength on semi-built-up crankshaft made of cast steel solely to
the pin fillet surface of semi-built crankshafts under estimate the effect of the compressive residual stress,
working conditions. applied by cold rolling, on fatigue strength.

Introduction 1. Summary of conventional results

Kobe Steel manufactures crankshafts for vessel 1.1 Evaluation by small-scale cold rolling test 1)
diesel engines as one of the company's flagship
products. In crankshafts, the fillets are subjected to Fig. 1 depicts a small-scale cold-rolling test. Each
maximum stress. Aiming at improving the fatigue specimen is made of a steel, "Throw Grade 3", which
strength of these parts, Kobe Steel developed a is one of the typical cast steel materials Kobe Steel
cold rolling technique. Not involving heating, this uses for semi-built-up crankshafts. Table 1 shows
technique does not cause thermal deformation and is the chemical composition and Table 2 shows the
superior in dimensional accuracy and productivity. It mechanical properties of the steel. A U-notched
was first adopted for a semi-built-up crankshaft portion is provided on the φ10mm specimen, and
made of cast steel in 1970 and has been adopted for cold rolling was applied to the notched portion. The
some solid crankshafts since 1995. Nowadays, this cold-rolled specimen was subjected to a rotary bending
technique has become one of the technologies fatigue test. The results are shown in Fig. 2. In this
making this company stand out among others. figure, the horizontal axis represents "mean stress,"
Cold rolling is a method for improving the fatigue while the vertical axis represents "fatigue strength."
strength of the fillets by the combined effect of
increased hardness caused by work hardening and 180
compressive residual stress applied by cold rolling. R2
φ10

φ8

Because hardness correlates well with fatigue

φ15

strength, the effect of hardness increase can be

predicted with relative ease. On the other hand, the (a) 10mm sized specimens
Load
effect of the compressive residual stress depends on
Work roll
the stress conditions of the fillets. The crankpin fillets
5
are under a combined load of bending and torsion. φ1
R1.95
Thus, in order to precisely evaluate the effect that the
compressive residual stress applied by cold rolling
has on the improvement of the fatigue strength, a Backup rolls
new evaluation method must be developed, taking (b) Small scaled cold rolling test
the combined loading into consideration. Fig. 1 Outline of small scaled cold rolling test using φ10mm
The purpose of the present study is to make sized specimens

7 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

 Table 1 Chemical composition of "Throw Grade3" 1,310 1,600

C％ Si％ Mn％ Cr％ Mo％ Ni％ V％

Spec.
≤ 0.50 ≤ 0.30 ≤ 0.50
＊ ＊ ＊ ＊
0.25 0.35 1.0 0.08
"Throw Grade3"
＊
nominal

1,330

2,674
Table 2 Material property of "Throw Grade3"
Tensile Reduction

φ1,090
Yield point Elongation
Specimen strength of area
（MPa） （％）
（MPa） （％）
Spec.
≧530 ≧310 ≧20 ≧45
"Throw Grade3"

Fig. 3 Dimensions of K98MC throw

600
Modified Goodman Diagram
U notched
by surface work hardening
500 (rolled)
Fatigue strength (MPa)

σw
by surface
400 σθ
work hardening
σr
300

σB
200 by surface S
work hardening 30deg.
100 U notched
(non-rolled) σB
60deg.
0
−400 −200 0 200 400 600 800 1,000 30deg. Pin
Mean stressσr (MPa)
0deg.
Fig. 2 Effects of surface work hardening and residual stress
on fatigue strength of φ10mm sized specimens after ：Measurement points Load
cold rolling S
Surface
Also included in the figure is a Modified Goodman Fig. 4 Residual stress measurement points
Diagram based on the fatigue strength, σw, and
tensile strength, σB, which reflects an increase in 0
hardness. The result indicates that, by regarding the
compressive residual stress as the mean stress, the Measurementσr
−100
Residual stress (MPa)

Measurementσθ
fatigue strength of the cold rolled specimen matches
well with the Modified Goodman Diagram that −200
accounts for the increase in hardness. It is concluded
from the above that the main factors improving the −300

fatigue strength of the steel, "Throw Grade 3", after

−400
cold rolling, are an increase in hardness caused by
work hardening and compressive residual stress
−500
applied by cold rolling. It should also be noted that 0 15 30 45 60 75
the compressive residual stress can be regarded as Pin fillet angle (deg.)

the mean stress. Fig. 5 Measured residual stresses on pin fillet surface

1.2 Residual stress distribution in actual surface in both the radial and circumferential
crankshafts 2) directions.

Kobe Steel produces large semi-built-up crankshafts, 1.3 Methodologies for predicting fatigue strength
including K98MC with a cylinder diameter of 980mm. considering combined loading
A real-scale throw was prepared using Throw Grade
3 steel, and residual stress was measured after A unified rule of the International Association
cold rolling. Fig. 3 depicts the K98MC throw. The of Classification Societies for crankshafts for diesel
measurement points for residual stress are shown in engines (IACS UR M53, hereinafter simply referred
Fig. 4 and the measurement results are shown in to as "IACS rule") 3) adopts an evaluation formula
Fig. 5. As shown in Fig. 5, a compressive residual that takes both bending and torsional stresses into
stress of about 400MPa is applied on the pin fillet account. However, it should be noted that the IACS

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 8

 rule focuses only on the stress amplitude of varying inclined at an angle θ; and
stress and disregards the effect of mean stress. To σθ is the normal stress amplitude in the plane
resolve this issue, the present study adopts a method inclined at an angle θ. It is to be noted that k , f are
of modifying bending stress amplitude according constants determined by the material and can be
to the modified Goodman diagram formula, as obtained from two fatigue strengths under different
previously reported.2) Formula (1) is the modified loads. The present study adopts Formula (5), which
Goodman diagram formula, while Formula (2) is an comprises fatigue strengths under alternating and
evaluation formula including a partial modification pulsating axial loads.
of the IACS rule. The effect of mean stress on σw, R＝0 k＋ 1＋k2 k＋ 1＋k2 σ
＝ , f＝ w, R＝−1 …… (5)
torsional stress is known to be negligible and thus is σw, R＝−1 2k＋ 1＋4k 2
2
ignored. This paper takes up the above three methods for
σrm evaluating fatigue strength under combined loading.
σ'ra＝σra 1−
σB …………………………… (1)
These methods were adopted for the fatigue test
σ'ra2＋3τa2 ＝σw ………………………………… (2) results and used for evaluating an actual crankpin
wherein fillet. Hereafter, the evaluation method using Formula
σra is the amplitude of bending stress working on (1) and Formula (2) is referred to as "Modified IACS",
the fillet surface the method using formula (3) is referred to as "Sines",
σrm is the mean stress working on the fillet surface and the one using Formula (4) and Formula (5) is
τa is the amplitude of torsional stress working on referred to as "Findley".
the fillet surface
σB is the tensile strength of the material, and 2. Validity of methods for predicting fatigue
σw is the axial load, or the rotary bending fatigue strength based on fatigue test
strength of the material
As a method for evaluating the fatigue strength 2.1 Material and specimen
under combined loading with a mean stress, Sines
proposed Formula 4) based on Formula (3) introduced Axial load and torsional load were simultaneously
from the octahedral shear stress criterion, applied in coordinate phase to each specimen made
1 2 2 2 of "Throw Grade 3 steel" during the fatigue tests.
2 （σax−σay）＋（σay−σaz）＋（σaz−σax）＋6（τ axy＋τ ayz＋τ axz）
2 2 2
Meanwhile, the material constants of three methods
                     
＝A−B（σmx＋σmy＋σmz）  (3) for evaluating fatigue strength were determined
σw, R＝−1 from fatigue strengths independently obtained with
A＝σw, R＝−1, B＝ −1
σw, R＝0 either an axial or torsional load respectively. Fig. 6
  
wherein depicts the shape of the specimen used for the test.
σai (i＝x, y, z) is the stress amplitude in x, y, z
directions, respectively;
σmi (i＝x, y, z) is the mean stress in x, y, z
directions, respectively;
τaij (i, j＝x, y, z) is the shear stress amplitude on
φ10

x, y, z surfaces, respectively;
σw, R＝−1 represents axial load fatigue strength
under alternating load condition; and 20
σw, R＝0 represents axial load fatigue strength
under pulsating load condition. 85.5
As one of the methods for evaluating fatigue (a) Axial fatigue test
strength under conditions where phase differences
exist in the combined loading, Findley used a critical
φ30

φ23

plane approach to propose (Formula (4))5). He

postulates that fatigue life is controlled by a
20

combination of alternating shear stress and maximum

normal stress on a critical plane at an angle θ, which
is determined by the condition that maximizes the 64 16
left-hand side of Formula (4).
（τθ＋kσθ） max＝f ………………………………… (4) 160.5

wherein (b) Torsional or combined loading fatigue test

τθ is the shear stress amplitude in the plane Fig. 6 Dimensions of fatigue test specimen

9 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

2.2 Testing conditions and results stress was applied in a pulsating manner. In the
condition with a compressive mean stress, a constant
Due to the limitation in the number of specimens, mean stress of -200MPa was applied axially to
each fatigue test was started with a low load, and the simulate the compressive residual stress applied by
load was increased step-by-step until failure. The cold rolling. The results are summarized in Table 3.
number of cycles for a given loading condition was
set to 3×106 cycles for fatigue tests under axial 2.3 Comparing methods for evaluating fatigue
loading, and 1×106 cycles for the fatigue tests under strength
torsional and combined loading.
In order to assess the effect of mean stress, axial The above three methods for fatigue strength
load fatigue tests were conducted under two evaluation were used to evaluate the results of the
conditions of alternating and pulsating loading. The fatigue tests conducted under combined loading.
effect of mean stress on the torsional fatigue strength Fig. 8 compares the evaluation results obtained by
was regarded as negligible and, thus, the condition these methods. Each vertical axis represents the
under which each torsional fatigue test was quotient of the value on the right-hand side divided
conducted did not include mean stress. The results of by the value on the left, for the respective evaluation
the axial loading and torsional fatigue tests are formulas. The value of the right-hand side was
shown in Fig. 7. The axial load fatigue strengths for calculated using the stress amplitude for the fatigue
reversed and pulsating stresses agree well with the test results in the uniaxial direction shown in Fig. 7.
modified Goodman diagram. The right-hand value is intrinsic to the material, and
The fatigue tests under combined loads were the fatigue limit is provided by the condition where
conducted under six conditions, as shown in Table 3. the value of the left-hand side matches that of the
In each test, both axial-stress and torsional stress right.
were applied simultaneously under three different An accurate fatigue strength evaluation requires
levels of mean stress, i.e., "without mean stress (zero the vertical axis value of Fig. 8 to be smaller than 1.0,
mean stress)," "with tensile mean stress" and "with at least for the case in which failure occurs. If the
compressive mean stress." The axial stress was value of the vertical axis becomes smaller than 1.0 for
applied in an alternating manner, while the torsional a non-failure (run out) condition, the evaluation is
regarded to be on the safe side. According to Fig. 8,
400 Axial fatigue test results (runout) all the methods yielded evaluation on the safe side
Axial fatigue test results (failure)
Torsional fatigue test results (runout)
Stress amplitudeσa (MPa)

300 Torsional fatigue test results (failure) Table 3 Fatigue test results under combined loading
modified Goodman diagram
Axial-torsion Axial stress Torsional stress
Result
200 fatigue tests Amplitude Mean Amplitude Mean
R＝−1.0

without mean 143.8 0.0 115.0 0.0 runout

R＝0.0
stress 149.9 0.0 120.0 0.0 failure
100

with tensile mean 130.0 130.0 65.0 65.0 runout

stress 140.0 140.0 70.0 70.0 failure
0
−200 −100 0 100 200 300 400 500 600
with compressive 200.0 −200.0 100.0 100.0 runout
Mean stressσm (MPa)
mean stress 210.0 −200.0 105.0 105.0 failure
Fig. 7 Results of Axial and torsional fatigue test

1.4 1.4 1.4

1.2 1.2 1.2

1.0 1.0 1.0

0.8
R（※1）

R（※1）

R（※1）

0.8 0.8
0.6 0.6 0.6

0.4 0.4 0.4 runout

runout runout
0.2 0.2 0.2 failure (※1)
failure failure
R＝(RHS value)/(LHS value)
0.0 0.0 0.0   in each equations
Modified

Findley
Modified

Modified
Findley

Findley

 Modified IACS using equation (2)

Sines
IACS
Sines

Sines
IACS

IACS

 Sines using equation (3)

 Findley using equation (4)

(a) without mean stresses (b) with tensile mean stresses (c) with compressive mean stresses
Fig. 8 Comparison of criterions under combined loading

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 10

 for conditions of alternating and pulsating loading; Table 4 Particular of studied engine
however, for the conditions with compressive mean MITSUI-MAN B&W
Type
stress, "Sines" evaluation yielded a result significantly 8K98MC-C
on the unsafe side. The "Modified IACS" and Output kW 45,680 (62,080HP)
"Findley" methods yielded evaluations on the safe MCR rpm 104.0
side regardless of the mean stress condition. The Cylinder number 8
"Modified IACS" method yielded the most accurate Cylinder bore mm 980
result among the three. Stroke mm 2,400
Firing order 1-8-3-4-7-2-5-6
3. Predicting effect of cold rolling on fillets Throw material Grade 3 (Cast steel)

In a manner similar to that previously reported,2) 100

a MITSUI-MAN B&W 8K98MC-C engine was used τ
80
σr
for operational stress analysis to determine the stress
60 σθ
generated on its pin fillet under working conditions.

Stress (MPa)
Table 4 summarizes the data for the engine. Fig. 9 40

shows the stress history of a pin-fillet with a fillet 20

angle of 30 degrees (the angle as defined in Fig. 4).
0
The pin-fillet was selected from the No.8 cylinder,
−20
since it is subjected to the maximum stress among
all the fillets. The figure indicates that there is −40
0 90 180 270 360
substantially no difference in the phase among the Crank angle (deg.)
three stress components; and the stress, σθ, in the Fig. 9 Surface stress history of pin fillet （No.8cyl. pin fillet
circumferential direction of the pin is smaller, in angle＝30deg.）
terms of both the stress amplitude and mean stress,
compared with either the radial direction stress, σr , 6.0

or the shear stress, τ.

5.0
The "Modified IACS" and "Findley" methods both
yielded safe-side results as shown in Fig. 8 and were 4.0
adopted for the evaluation of the stress generated on
R（※2）

the pin fillet surface of the No.8 cylinder under 3.0

working conditions. Fig.10 shows the evaluation
2.0 Modified IACS withoutσθ
results for the fatigue strength of the pin fillet
Findley withoutσθ
surface under combined loading. The effect of cold
1.0 Findley (considering all stresses)
rolling has not been taken into account in this
figure. The "Modified IACS" method cannot account 0.0
−15 0 15 30 45 60 75
for the stress, σθ, in the circumferential direction of Pin fillet angle (deg.)
the pin. Because of this, the "Findley" method was (※2)
R＝(RHS value)/(LHS value) in each equations
used to confirm the effect of the stress in the  Modified IACS using equation (2)
circumferential direction of the pin, and it was  Findley using equation (4)

confirmed that the effect is negligible. Both the Fig.10 Evaluation of fatigue strength under combined loading
evaluations indicate that the stress currently of pin fillet (No.8cyl.)
generated is no greater than the fatigue limit. The
"Findley" method yielded a result significantly more between the bending stress and torsion stress was
on the safe side than did the "Modified IACS." fixed at 1/2, which is same as in the case of the ratio
Also confirmed was the effect of the compressive of maximum stress under the working conditions
residual stress applied by cold rolling on the fatigue shown in Fig. 9. The compressive residual stress
strength of the pin fillet surface. Either in Formula (2) applied by cold rolling was obtained from the actual
of the "Modified IACS", or in Formula (4) of the measurements shown in Fig. 5. Fig. 11 shows the
"Findley," the permissible stress is determined by permissible stresses for the pin fillet surface
making the value of the right-hand side equal to that determined by the "Modified IACS" and "Findley"
of the left-hand side, on the assumption that the evaluations, as well as the way in which the
stress conditions on the pin fillet surface were fixed. permissible stresses were affected by the compressive
In this paper the stress condition on the pin fillet residual stresses applied by cold rolling. This figure
surface was set to a pulsating stress, and the ratio indicates that Findley's evaluation yields results

11 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

 250 1.6

1.4

Permissible stress (MPa)

200
1.2

Rate of improvement
150 1.0

0.8
100 0.6
Modified IACS with cold rolling
Modified IACS Modified IACS
0.4
50 Findley with cold rolling Findley
Findleyy 0.2

0 0.0
0 15 30 45 60 75 0 15 30 45 60 75
Pin fillet angle (deg.) Pin fillet angle (deg.)
(a)Permissible stress (b) Rate of improvement

Fig.11 Rate of improvement of pin fillet permissible stress by cold rolling

more on the safe side than is the case with the mean stress. The validity of three methods for
Modified IACS; however, both methods evaluate the determining fatigue strength under combined
effect of the compressive residual stress on the loading was evaluated. As a result, it was found
permissible stress as being almost the same. It was that the "Modified IACS" and "Findley" methods
also found that, on the pin fillet surface, the effect provide evaluation that is on the safe side, regardless
of the compressive residual stress applied by cold of the mean stress. The "Modified IACS" has turned
rolling improves the permissible stress by out to be a very convenient and effective evaluation
approximately 40％. method in cases where the combined stress consists
As reported in this paper, both the "Modified only of bending and torsion without a phase
IACS" and "Findley" yield almost the same results for difference.
the evaluation of compressive residual stress applied This evaluation method was adopted for the pin
by cold rolling on the fatigue strength of fillet surfaces. fillet surface of an operating engine. It turned out that
Findley's method makes it possible to account for all the compressive residual stress applied by cold
the stress components and phase differences; however, rolling improves the permissible stress on the pin
it requires repeated computation to determine the fillet surface by approximately 40％. This study does
conditions that maximize the left-hand side of not account for the effect of work hardening caused
Formula (4). In the stress condition described in this by cold rolling. Thus, a more significant improvement
paper, the stress in the circumferential direction of in fatigue strength can be expected for a material
the pin can be regarded as negligible, and almost having a greater work-hardening effect. The cold
no phase difference exists. This has made accurate rolling technique has been implemented in practice
evaluation by the "Modified IACS" possible. The and is expected to see more use in the future.
"Modified IACS" is a very convenient and effective
evaluation method under stress conditions where References
only bending and torsion are combined, without a
phase difference. 1) T. Hamada, et al., Annual Spring Meeting (131st), Iron and
Steel Institute of Japan, p.394.
2) M. Matsuda et al., R&D KOBE STEEL ENGINEERING
Conclusions REPORTS, Vol. 59, No. 1 (2009), pp.89-93.
3) IACS UR M53, Calculation of Crankshafts for I. C. Engines.
Fatigue tests were conducted under combined 4) G. Sines, NACA Tech. Note, 3495 (1955).
loading by regarding the effect of cold rolling as 5) W. N. Findley, J. Eng. Ind., Nov. (1959), pp.301-306.

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 12

 Effect of Alloying Elements on Machinability and Hot
Workability of α-β Titanium Alloy Containing Fe and C
Shogo MURAKAMI＊1, Dr. Katsuhiko OZAKI＊1, Kousuke ONO＊2, Yoshio ITSUMI＊3
＊1
Materials Research Laboratory, Technical Development Group
＊2
Titanium Marketing & Technical Service Section, Titanium Division, Iron & Steel Business
＊3
Titanium Research & Development Section, Titanium Division, Iron & Steel Business

Many studies have been conducted on titanium alloys to This paper summarizes the results of the study on
improve their machinability and hot workability. They the effect of alloying elements on the machinability of
are inferior in machinability and hot workability when the KS EL-F alloy, as well as the results found for
compared with other structural materials, such as steel the hot workability and machinability of modified
and aluminum alloys. The KS EL-F (Ti-4.5Al-4Cr- compositions. Also discussed is the aging embrittlement
0.5Fe-0.15C) that has been developed has mechanical of the modified compositions, since the KS EL-F alloy
properties equivalent to those of Ti-6Al-4V at room is known to exhibit aging embrittlement 5) and has
temperature and has superior hot workability. It was limited applications where elevated temperatures
found that increasing Fe and decreasing Cr concentration are encountered.
in the chemical composition of KS EL-F improves both
machinability and hot workability. The modified alloys 1. Experimental procedure
(Cr: 1 to 2.5%, Fe: 1.25 to 2%) were found to decrease
tool tip wear by almost 30% and the hot deformation 1.1 Sample preparation
stress by about 10%. Furthermore, decreasing Cr
suppressed aging embrittlement caused by the To experimentally select an alloy system, eight
precipitation of TiCr2. alloys as shown in Table 1 were prepared; the KS
EL-F alloy (Ti-4.5Al-4Cr-0.5Fe-0.15C, hereinafter
Introduction referred to as "EL-F") being the reference composition;
six alloys containing Fe Mo, V, Ni, Mn, and Co, these
As typified by Ti-6Al-4V alloy, α-β type titanium elements being substitutes for Cr ; and one alloy
alloys have excellent specific strength, heat resistance containing Sn as a substitute for Al. The chemical
and corrosion resistance, and are often used for compositions were designed such that the Mo
parts for aircrafts, automobiles and motorcycles1). equivalent falls in the range from 6.25% to 6.35%
Titanium alloys are increasingly used for joint parts and Al equivalent in the range around 4.5%.
with carbon fiber reinforced plastic (CFRP), because Chromium was replaced to prevent the precipitation
they have a coefficient of thermal-expansion close of TiC and to suppress the production of aging
to that of CFRP and do not cause electrolytic precipitates, TiCr2. Aluminum was replaced because
corrosion. This application is now commonly the substitution was expected to further decrease the
found; for example, it was used in the fuselage of the deformation resistance at elevated temperatures. The
Boeing 787, which made its successful first flight in tested materials were melted by vacuum arc melting
November, 2009. The demand for titanium alloys is to prepare button ingots of about φ40×20mm in
envisaged to increase in the field of aviation2). The size, and each button ingot was forged into a size of
mechanical properties of α-β type titanium alloys about φ25×55mm. In order to clarify in more detail
can be adjusted by various heat treatments; however, the effect of alloying elements on machinability, it is
commonly-used Ti-6Al-4V alloy suffers from a high preferable to have a beta-annealed microstructure
machining cost, in addition to its high material cost, that consists of many alpha laths in beta matrix. To
due to its poor forgeability and machinability
compared with other materials such as steel and Table 1 Chemical compositions of samples for machinability
aluminum alloys. This has led to the development （mass％）
and use of various other alloys. KS EL-F is one of Mo Al
No. Al Cr Fe C Others
such alloys and has a significantly improved hot equivalent equivalent

workability 3). EL-F 4.5 4 0.5 0.15 − 6.25 4.5

5％Mo 4.5 − 0.5 0.15 Mo：5.0 6.25 4.5
The KS EL-F alloy contains carbon, which
7.5％V 4.5 − 0.5 0.15 V：7.5 6.25 4.5
decreases the alloy's deformation resistance at
2Fe-1Cr 4.5 1 2 0.15 − 6.25 4.5
elevated temperatures, while maintaining its strength 4％Ni 4.5 − 0.5 0.15 Ni：4.0 6.25 4.5
at room temperature. However, TiC may precipitate, 3％Mn 4.5 − 0.5 0.15 Mn：3.0 6.35 4.5
depending on thermo-mechanical treatment, and 3％Co 4.5 − 0.5 0.15 Co：3.0 6.35 4.5
the precipitates can increase tool wear 4). 7.5Sn-2Al 2 4 0.5 0.15 Sn：7.5 6.25 4.5

13 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

achieve this, all the compositions were annealed at a The specimens were heated to 875℃ or 810℃ in an
temperature of 1,100℃, which is higher than the beta electrical furnace, held at their respective temperatures
transus, held at the temperature for 2 hrs. and for 50min, and hot-pressed at 700℃ or 650℃,
subsequently air-cooled. respectively, by a 1,600tonne press. The upper die
Two compositions that were expected to be more was kept at a temperature ranging from 120 to 130℃
excellent on the machinability (in Table 2) were then and the lower die at a temperature ranging from 150
melted in a cold crucible induction melting (CCIM) to 170℃. The ends of each specimen were confined
furnace to prepare ingots of about φ155×165mm in during the pressing. The strain rate was controlled to
size. Each ingot was forged into a round bar with a 1 to 2/s. The specimens were then investigated for
diameter of φ60mm, and the bar was annealed at the existence or non-existence of cracking at the
800℃ for 1.5hrs. Each annealed bar was investigated bottom of the notches. A hot working simulator was
for its hot-workability, machinability and aging used to evaluate deformation resistance on standard
behavior. The samples of Ti-6Al-4V and "EL-F" alloys, samples, each having a dimension of φ8×12mm, at
both used as reference materials, were prepared from working temperatures in the range from 400 to
commercially available round bars of φ60mm. 800℃. The crosshead speed of the tester was set to
15mm/s, and the maximum deformation stress was
1.2 Evaluation method taken as the value for the deformation resistance.
To study aging behaviors, each annealed sample
The machinability of each composition was was reheated to a temperature ranging from 300 to
evaluated using an optical microscope to determine 700℃, held at that temperature for 8hrs and
the flank wear width of the cutting edge of a tool- subsequently air-cooled. Tensile tests were conducted
chip after cutting. The details of the cutting to study the change in strength and ductility. The
conditions are shown in each figure. tensile tests, according to JIS Z2241, were conducted
Microstructures were observed under an optical on test pieces, each having a parallel portion
microscope on each sample, which was mirror diameter of φ6mm and a gauge length of 30mm.
polished and etched by an etchant (i.e., water : nitric A database, Ti-DATA ver.2, was adapted for
acid : hydrofluoric acid= 80 : 15 : 1). Hardness was thermodynamic calculation software, Thermo-Calc
measured using a Vickers tester with a load of 10kg. ver.P, to determine the quasi-binary phase diagram
Hot workability was evaluated on specimens, as and the solid solubility limit of carbon.
shown in Fig. 1, each having a dimension of φ15×
22.5mm with notches 0.3mm deep on both sides. 2. Experimental results and discussion

Table 2 Chemical composition of high Fe system test alloys 2.1 Effect of alloying elements on machinability
Alloy Al Cr Fe C
EL-F 4.5 4 0.5 0.15 In general, tool wear increases with an increase in
F-1 4.5 2.5 1.25 0.15 the hardness of the work material. All of the
F-2 4.5 1 2 0.15 compositions shown in Table 1 have an almost
identical Vickers hardness of 310±10HV, hence the
effect of hardness is regarded as negligible in this
study. The microstructure of each composition
φ15.0mm

exhibits a β-anneal structure, as shown in Fig. 2, with

Position A
EL-F 5％Mo
22.5mm
（L／D＝1.5）

Position A
0.3mm

60° 25μm 25μm

2Fe-1Cr 3％Mn

R0.05

Deepness of notch：0.3mm
Angle of notch：60°
Tip radius of notch：0.05mm 25μm 25μm

Fig. 1 Test specimen on limit of compressibility Fig. 2 Microstructure of representative samples

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 14

 a small amount of TiC remaining. long as the carbon is added within the solid solubility
Dry cutting was conducted with the cutting limit shown in Table 3, because tool-wearing TiC
parameters as follows: cutting speed, 100m/min; precipitates near the beta-transus. Conversely, if the
feed rate, 0.1mm/rev; cutting depth, 0.5mm; and addition exceeds the solid solubility limit, the
cutting distance, 50m. The flank wear widths after precipitation of TiC must be inhibited by adjusting
cutting are shown in Fig. 3. Compared with the tool the conditions of hot-working and/or annealing,
wear by the EL-F reference composition, the including the annealing temperature and time. As
materials designated as 5% Mo and 2Fe-1Cr show shown in Table 3, the solid solubility limits of carbon
favorably small wear volumes. On the other hand, for the 5% Mo material and for 2Fe-1Cr material are
the materials designated as 7.5% V and 4% Ni show
a tool wear volume almost the same as that of EL-F,
and the ones designated as 3% Mn, 3 Co and 7.5Sn- 100 V＝50m/min
KS EL-F V＝100m/min
2Al show wear volumes greater than that of the
reference material. 80

Frank wear width (μm)

As shown in Fig. 2, the 5% Mo and 2Fe-1Cr
60
materials, each exhibiting a small wear volume, have
alpha laths that are slightly larger. In general, larger
40
strain concentration in a secondary shear zone Tool-tip：Carbide tip (SANDVIK H13A )
facilitates the generation of cutting chips, improving Feed：0.1mm/rev Depth of cut：0.5mm
20 Lubrication: none (dry)
machinability 6). Because large alpha laths are Cutting distance：100m
considered to inhibit the strain concentration, the 0
structure containing the phase was originally 0.0 0.5 1.0 1.5 2.0 2.5
Volume fraction of TiC (％)
regarded to be unfavorable for machinability.
Nevertheless, the structure exhibits a smaller tool Fig. 4 Influence of TiC volume fraction on frank wear of
cutting tool tip4)
wear volume, indicating that the substitution of Cr
with Fe and/or Mo effectively reduces tool wear.
On the other hand, a large volume fraction of TiC
EL-F ：Ti-4.5Al-4Cr-0.5Fe-0.15C
precipitates increases tool wear, as shown in Fig. 4 4);
2Fe-1Cr ：Ti-4.5Al-1Cr-2.0Fe-0.15C
hence, it is preferable to make the solid solubility
1,300
limit of carbon as large as possible when designing
a composition. Using the thermodynamic calculation 1,200

software, Thermo-Calc, the solid solubility limit of β

1,100 β＋TiC
Temperature (℃)

carbon near the beta transus was calculated for each 1,000
composition of 5% Mo, 7.5% V, 1Cr-2Fe, 4% Ni and
900
7.5 Sn-2Al. The results are shown in Table 3. Fig. 5
depicts a quasi-binary phase diagram in which the 800 α＋β
horizontal axis represents the carbon concentration 700
of EL-F and 2Fe-1Cr materials. The addition of carbon 600 α＋β＋TiC
should not noticeably deteriorate machinability as
500
0 0.1 0.2 0.3 0.4 0.5
Concentration of C (mass％)
Tool-tip：Carbide tip (SANDVIK H13A )
Cutting speed：100m/min Fig. 5 Calculated phase diagram of Ti-Al-Cr-Fe-C system
100 Feed：0.1mm/rev Depth of cut：0.5mm
90 Lubrication: none (dry) Cutting distance：50m
Frank wear width（μm）

80
Table 3 Solubility limit of C on beta transformed temperature
70
in test alloys
60
No. Solubility limit of C (mass％)
50
40 EL-F 0.140

30 5％Mo 0.182
20 7.5％V 0.133
10 2Fe-1Cr 0.198
0 4％Ni 0.118
EL-F

5％Mo

7.5％V

2Fe-1Cr

4％Ni

3％Mn

3％Co

7.5Sn-2Al

3％Mn no data
3％Co no data
Fig. 3 Frank wear of test alloys for machinability 7.5Sn-2Al 0.259

15 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

greater than for the base material, EL-F. The greater doubled cutting distance. The decreased wear can be
solubility limit is preferable for inhibiting the attributed to the microstructure, shown in Fig. 6, in
precipitation of TiC, eliminating the need for which coarse TiC precipitates do not exist, and
specifically adjusting the conditions for forging, or equiaxial α structure has been formed to be different
for annealing. from α lath structure shown in Fig. 2. In general, the
As described above, it has been found that equiaxial α structure has a fracture toughness lower
substituting Cr with Mo, or with Fe, effectively than that of the alpha lath8), which is considered to
improves the machinability of EL-F. The substitution facilitate the generation of cutting chips and to
with Fe is preferable because the element is more decrease tool wear.
widely found and less costly 7). The following focuses
on the compositions in which Cr is substituted with
Fe. KS EL-F F-1

2.2 Hot workability and machinability of α-β type

titanium alloys containing high amount of Fe
25μm 25μm
Among alloying elements, Fe segregates
significantly during actual production involving F-2
melting and casting. Taking this into consideration,
the compositions shown in Table 2 were prepared
for the evaluation of hot workability. Fig. 6 shows
the microstructures of the samples of these alloys.
25μm
The samples have almost the same Vickers hardness
of 340±10HV. Fig. 6 Microstructure of high Fe system test alloys
Compression tests were conducted with the ends
of each specimen confined to see if cracking occurred
in the samples. Fig. 7 shows the results of the 100 700℃
Limit of compressibility without crack (％)

650℃
evaluation. Compared with the conventional α-β 90

type titanium alloy, Ti-6Al-4V (hereinafter referred 80

: no crack
to as "Ti-64"), the EL-F and the compositions with 70
high content of iron, i.e., F-1 and F-2, have very high 60
crack limits. In particular, the high-Fe compositions 50
have crack limits even higher than that of EL-F. Also 40
studied was the deformation resistance of each alloy.
30
As shown in Fig. 8, the deformation resistance of
20
Ti-64 depends less on temperature, while both EL-F
10
and the high iron compositions (F-1, F-2) have
deformation resistance that decreases significantly at 0
Ti-64 EL-F F-1 F-2
elevated temperatures. At 600℃, F-1 and F-2 have a Fig. 7 Limit of compressibility of conventional and high Fe
deformation resistance almost 10% smaller than that system alloys
of EL-F. At 600℃, Fe, in α phase, has a diffusion
coefficient approximately two orders of magnitude 1,000
Deformation maximum stress (MPa)

higher than that of Cr and, in β phase, three to four 900

Ti-64
EL-F
orders of magnitude higher than the same. Hence F-1
800
it is presumed that the decreased Cr and increased F-2

Fe content promote diffusion, further decreasing 700

deformation resistance. 600

Fig. 9 compares tool wear, wherein each sample 500
was cut at various cutting speeds ranging from 10 to 400
100m/min for a distance of 100m. Compared either
300
with Ti-64, or with EL-F alloy, the alloys with higher
Fe content wear the tools much less and exhibit the 200
300 400 500 600 700 800 900
most favorable machinability. Comparing them with Deformation temperaure (℃)
the results shown in Fig. 4, they exhibit smaller tool Fig. 8 Deformation stress of conventional and high Fe
wear, despite the same cutting conditions and the system alloys

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 16

 50
Ti-64 to occur in the EL-F alloy by the precipitation of
EL-F
45 F-1 TiCr26), and the decreased Cr content in F-1 and F-2
F-2
Frank wear width (μm)

40 is considered to inhibit the aging embrittlement. It

35 should be noted that the F-1 and F-2 alloys have a
30
slight difference in the temperature range where the
25
20 Tool-tip：Carbide changes in strength and ductility occur. This may be
15 tip(SANDVIK/H13A) attributable to other precipitates, such as TiFe and
Feed：0.1mm/rev
10
Depth of cut：0.5mm TiCr2, as well as to TiC, the details of which are to be
5 Lubrication：none(dry) the subject of future study.
0
0 20 40 60 80 100 120
Cutting speed (m/min) Conclusions
Fig. 9 Frank wear of conventional and high Fe system alloys
A study was conducted on machinability and hot
YS_EL-F TS_EL-F workability, using KS EL-F alloy as a reference. The
YS_F-1 TS_F-1 results are as follows:
1,200 YS_F-2 TS_F-2 1) Using the KS EL-F composition (Ti-4.5Al-4Cr-
1,150 0.5Fe-0.15C) as a base, alloys were prepared with
Yield/Tensil strength (MPa)

1,100 decreased Cr and Al content. These elements

1,050 were substituted with other alloying elements
1,000 (i.e., Mo, V, Fe, Ni, Mn, Co, Sn). Each alloy was
950
evaluated for its machinability. It has turned
out that substituting Cr with Mo and/or Fe
900
effectively improves the machinability.
850
Aging time：8h 2) In an equiaxial α structure, the alloys (1 to 2.5Cr-
800 1.25 to 2Fe), which has a higher Fe content
200 300 400 500 600 700
Aging temperature (℃) compared with Ti-6Al-4V, or with KS EL-F,
Fig.10 Strength after aging on high Fe system alloys (KS exhibit tool wear decreased by about 30%. The
EL-F data : ref. 5)) alloys have an approximately 10% lower hot
deformation resistance at 600℃ and crack limit
EL_EL-F RA_EL-F
compression rates no less than 5% higher.
60
EL_F-1 RA_F-1 3) Thermodynamic equilibrium calculation shows
Elongation /Reduction of area (％)

EL_F-2 RA_F-2
that the alloys containing high Fe have large solid
50 solubility limits of carbon near the beta-transus.
40 This makes the alloys less prone, compared with
KS EL-F, to the precipitation of coarse TiC, which
30
can adversely affect the machinability.
20 4) The alloys with compositions of decreased Cr
and increased Fe have limited precipitation of
10
Aging time：8h TiCr2, which prevents the aging embrittlement
0
200 300 400 500 600 700
that occurs in KS EL at around 500℃.
Aging temperature (℃) A study is currently in progress to commercialize
Fig.11 Ductility after aging on high Fe system alloys (KS EL-F a modified alloy, KS 531C, which reflects the
data : ref. 5)) knowledge described in this paper. This work has
been conducted as a part of the project, "Aerospace
Industry Innovation Program-Advanced Materials &
2.3 Aging behavior of α-β type titanium alloy Process Development for Next-Generation Aircraft
having high Fe content Structures" under the contract with RIMCOF
Research Center of Materials Process Technology
The EL-F alloy and high-Fe alloys shown in Center, founded by Ministry of Economy, Trade and
Table 2 were aged and subjected to tensile testing. Industry (METI) of Japan. We would like to express
Fig.10 and Fig.11 show their tensile properties after our sincere gratitude to all who were involved in this
the aging. Each alloy has a high strength at around project.
500℃ with decreased ductility. Compared with the
EL-F, the F-1 and F-2 exhibit a much smaller
reduction of ductility. Aging embrittlement is known

17 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

References 4) K. Ozaki, R&D KOBE STEEL ENGINEERING REPORTS,
Vol.55, No.3 (2005) pp.61-63.
1) ADVANCED MATERIALS & PROCESSES/OCTOBER, (2003), 5) K. Ono et al., KEIKINZOKU, Vol. 55, No. 12 (2005), pp.673-
p.38. 677.
2) "Material Technologies supporting the development of 6) Recht. R. F., Trans. ASME (1964), pp.186-193.
aircrafts (Japanese)," Bulletin of The Iron and Steel Institute of 7) M. Niinomi, 61st Shiraishi Kinen Koza, (2009), pp. 1-15.
Japan (Ferrum), Vol.11, No.2 (2006), pp.2-6. 8) M. Niinomi et al., Tetsu-to-Hagane, Vol. 72, No. 6 (1986),
3) H. Oyama et al., Materials Science Forum, 426-432 (2003) p.713. pp.99-106.

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 18

 Pre-coated Titanium Sheet with Excellent Press
Formability
Akihisa FUJITA＊1, Yoshio ITSUMI＊1, Tadashige NAKAMOTO＊2, Kayo YAMAMOTO＊2
＊1
Titanium Research & Development Section, Titanium Division, Iron & Steel Business
＊2
Sheet Products Development Department, Research & Development Laboratory, Iron & Steel Business

Plate type Heat Exchangers (PHEs) are one of the main operated at higher pressures.
applications for commercially pure titanium. The plates, Conventional PHEs use the softest grade of
each consisting of a titanium sheet press-formed into a commercially pure (CP) titanium. If PHEs can
complex corrugated pattern, are directly linked to high employ harder CP titanium while preserving the
performance such as that seen in the thermal favorable formability of the soft material, the
conductivity of the PHE. Now Kobe Steel has newly application of titanium will expand to higher
developed a pre-coated titanium sheet with excellent pressure applications with possible weight reduction
press-formability, whose design includes a lubricant achieved by wall thinning. Thus, a material for PHE
coating that is easily removed by alkaline cleansing. plates is required to have both favorable press
formability and high strength.
Introduction The strength and formability of a material,
however, are in a trade-off relationship. Thus, Kobe
Plate type heat exchangers (hereinafter referred Steel first focused on surface lubrication as a means
to as "PHEs"), which exploit large amounts of sea for improving formability. Various methods have
water as their cooling media, are used in industrial been known for such lubrication. One common
fields including chemical plants, power-generating approach is to attach a lubricating film to the surface
facilities and large transport vessels. PHEs employ to be formed2); however, this method costs time and
large amounts of titanium for their primary parts money for attaching and detaching the film in actual
such as their plates and piping because titanium operation. Another approach is to oxidize, or to
exhibits excellent resistance against seawater nitride, the surface3); however, this method suffers
corrosion. Now PHEs have become one of the from hardness and brittleness in the resulting surface
main applications of titanium1). Fig. 1 illustrates layer, making the layer prone to crack in press
the operating principle of a PHE. A PHE is an patterns that are difficult to form. It also suffers from
apparatus for exchanging heat between hot water the fact that the surface layer may inhibit heat
and seawater, each flowing in the opposite transfer. Taking these facts into account, Kobe Steel
direction, wherein the flows are separated by developed a method of pre-coating a lubricant, such
titanium plates. Heat exchange occurs directly via that the pre-coated layer can be removed easily
these titanium plates. Each plate consists of a during the alkali cleaning that follows press
titanium sheet which is formed into a complicated forming. This method has enabled the use of JIS
corrugated shape and determines the heat transfer Class-2 CP titanium with a higher strength than that
performance of the heat exchangers. The plate of conventional JIS Class-1 CP titanium, with the
materials are also required to have increased conventional press formability.
strength, as the PHEs are increasingly being This paper introduces an adequate evaluation
method uniquely developed for numerically
evaluating the press formability of the titanium
sheets used for PHE plates. Also introduced is a
titanium sheet pre-coated with lubricant, which was
developed using this evaluation method.

Low 1. Kobe Steel's approach to evaluating press

temperature
High
formability
temperature
During press forming, cracking occurs more
frequently at sites called "flow passages" or "gaskets".
The deformation mode at these sites is close to the
mode of biaxial stretch forming4), 5). Thus, the Eriksen
value is used as a simple index of press formability.
Fig. 1 Principle of plate type heat exchanger In reality, however, not only stretching, but also

19 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

other deformation modes, such as deep drawing, wherein E is the score at each measuring point and
come into play together. Thus, a method must be R is the ridge radius (mm) at each measuring point.
established for quantitatively evaluating press No cracking at any of the sites yields 100%, while the
formability in real-life situations. cracking at all the sites yields 0%. The reason for
To achieve this, attempts have been made to dividing by R is to account for the effect of bending
quantify press formability by scoring the conditions of deformation at each ridge with radius R.
cracking and necking after pressing, according to
the following methods. For simplified evaluation, 1.2 Evaluation using large die
pressing was conducted using a compact test die
which simulates a fishbone pattern, also known as The shape of the large die is as follows:
the herringbone pattern, which is commonly found ・Size : 500mm square; (forming portion : 300mm
in the heat transfer portion of PHE plates. Also used square)
was a large die covering a size close to that of an ・Pitch between ridges : 14.9mm
actual PHE. ・Ridge height : 8.1mm
It is to be noted that there are still issues ・Radius of the ridges R＝3.4 (mm)
remaining, including the fact that the present press- The test samples were sized to 500mm square and
formability evaluation does not correlate sufficiently were pressed by a 1,000tonne servo-press. The
well with actual production. Further study will be criterion used for the compact die was adapted,
conducted to improve the evaluation accuracy. making the forming height of the ridge portions

1.1 Evaluation using compact die

A
B C
The following describes the shape of the compact R
die. 0.6
160mm 1.4
・Size: 160mm square (forming portion: 100mm 1.0
0.8
square) 1.8 A'
・Pitch between ridges: 10.0mm 0.4
・Ridge height: 4.0mm B' C'
・Radii of ridges R : 0.4, 0.6, 0.8, 1.0, 1.4, 1.8 (mm)
・Number of ridges: 6 100mm
The test samples were sized to 160mm square and 160mm
were pressed by an 80tonne hydraulic press. The A-A' Cross section
Convex
3.4mm
ridges were formed with a height of 3.4mm, the A A'
height having been determined to exhibit clear Concave
differences in cracking conditions between JIS Class- 10.0mm

1 and Class-2 titaniums. Fig. 2 Pressed test shape and positions for scoring in order
After test pressing, the samples were quantitatively to evaluate the pressformability using compact size
test die
ranked by the existence or non-existence of cracks
and/or neckings at predetermined sites, as follows. A
In a press-formed sample (Fig. 2), a total of 24 sites
were selected for evaluation, i.e., the 18 convex
portions represented by the crossing between the 6
ridge-lines, the sites of potential cracks, and the three
500mm

broken lines (A-A', B-B', C-C'), and the 6 concave

portions on the broken line A-A', the concave
portions where cracks can also originate. The
following scoring was applied for each site:
A'
・No cracking (soundness) : score 4
300mm
・Slight necking : score 3
500mm
・Significant necking : score 2 A-A' Cross section
Convex
・Small cracking : score 1 4.5mm
A A'
・Large cracking : score 0
Concave
Based on these scores, the overall score was 14.9mm
calculated as follows: Fig. 3 Pressed test shape and positions for scoring in order
  Score (%)＝(Σ(E/R))/(Σ(4/R))×100 ………(1) to evaluate the pressformability using large size test die

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 20

 4.5mm. There are a total of 92 evaluation sites in the well to the titanium sheet, sufficiently follows the
areas surrounded by the broken lines in Fig. 3, deformation and can easily be removed by alkali.
including ridge ends and ridge central regions Colloidal silica hardens the coating. Polyolefin wax
(convex portions and concave portions). Scoring was improves lubricity against dies and was added to
conducted in the same manner as in the case of the reduce the kinetic coefficient of friction of the pre-
compact die. It is to be noted that R is set to be coating.
constant to form a shape close to that of a real To study press formability under different
product. lubricity, three types of lubricated titanium sheets
  Score(%)＝ΣE/(4×92)×100 …………………(2) were prepared using JIS Class-2 CP titanium, which
has properties as shown in Table 2. One type of
2. Features of lubricant pre-coated titanium sheets sheet was coated with press oil (SUNPRESS S-304
manufactured by Sugimura Chemical Industrial Co.,
To study the difference in deformation between Ltd.), another was covered with a film having the
film lubrication and press oil lubrication, samples composition of the pre-coated lubricant without
were pressed by the compact die with a lubricant film polyolefin wax (i.e., acrylic resin＋colloidal silica),
and with a press oil (SUNPRESS S-304 manufactured and the remaining one was pre-coated with
by Sugimura Chemical Industrial Co., Ltd.). Fig. 4 lubricant. The coatings were applied with a roll
compares the cross-sections of the ridge lines (R＝0.8, coater and were dried in a thermostatic oven into a
1.0, 1.8) of the pressed samples. In the area thickness of 1.0μm.
surrounded by dotted circles, the sheet thickness The coating thickness was adjusted by varying
is decreased in relation to its periphery. The the solid content in the pre-coat solution. An X-ray
comparison indicates that film lubrication results in fluorescence analyzer (MIF-2100, manufactured by
less cracking and necking compared with press oil Shimadzu Corporation) was used to quantify the
lubrication. In the case of press oil lubrication, the die amount of Si in the film. The coating thickness (μm)
comes into contact locally with the titanium sheet as was determined from the coating quantity (g/m2),
the forming proceeds, restraining and immobilizing obtained by Formula (3), which was converted by
the sheet. In the meantime, the forming continues as Formula (4).
a whole, causing necking and eventual cracking. On   Coating quantity(g/mm2)＝
the other hand, in the case of film lubrication, the die Si×60×100/28×C×1000 … (3)
does not directly contact the titanium sheet as long as wherein Si represents the amount of Si in the
the film is not broken, which assures the fluidity of coating(mg/m2), C represents the concentration of
the sheet, allowing the supply of the material from SiO2 in the surface treatment composition(%), the
the surrounding region. This is considered to be numeral 28 is the atomic weight of Si, and the
the reason for the resulting uniform deformation numeral 60 is the molecular weight of SiO2.
compared with the case of press oil lubrication. Thus, Coating thickness (μm)＝(Coating quantity×
in order to obtain a high formability, it is important 0.1/2.2)＋(Coating quantity×0.9/1.0)… (4)
to reduce the friction force as much as possible. Formula (4) is applicable to a coating, that is 10%
To improve lubricity and, in particular, to reduce SiO2 (specific gravity 2.2) and 90% resin ＋ wax
the kinetic coefficient of friction, a lubricant pre- (specific gravity 1.0).
coated sheet was prepared. Table 1 shows the The press formability was quantified based on
composition of the lubricant pre-coating. Acrylic the method using the compact die as described in
resin was chosen as the base resin because it adheres Section 1.1. The kinetic coefficient of friction was

：Thickness decrease point Table 1 Composition of pre-coated layer

Press oil mass％ Contents
・Acrylic resin consisted of
2.0
R＝2.0 Resin 80
R＝2.0 2.0
R＝2.0  alkylacrylate-methacrylate copolymer
1.8
R＝1.8 R＝0.8 R＝1.0 Colloidal ・Amorphous SiO2
10
silica ・Alkali Na2O
Wax 10 ・Polyolefin wax
Polyethylene film

R＝2.0 2.0
R＝2.0 2.0
R＝2.0 Table 2 Tensile properties of JIS Class-2 titanium sheet
1.8
R＝1.8 R＝0.8 R＝1.0 0.2％Yield Tensile
Tensile Elongation
strength strength
direction （％）
Fig. 4 Comparison between cross sections of pressed （MPa） （MPa）
samples using press oil and polyethylene film as L direction 243 396 31
lubricant T direction 288 389 29

21 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

measured by a tester (HEIDON by Shinto-Science) 100

in the longitudinal direction with a test load of

Removal rate of pre-coated layer (％)

500g, a sliding speed of 100mm/min and a sliding 80
distance of 40mm. The partner material was a
stainless steel ball of φ10.0mm. 60
Fig. 5 shows the relationship between the kinetic
friction coefficient and press formability of the pre-
40
coated sheet that was prepared. It can be concluded,
despite the limited data, that the kinetic friction
coefficient of the material is decreased by the 20

application of the coating comprising resin and

colloidal silica and is further decreased by the 0
addition of polyolefin wax. This is accompanied by 0 30 60 90 120 150
Dipping time (s)
an improvement in formability.
Fig. 7 Removability of pre-coated layer by dipping in alkaline
To determine the optimum pre-coating thickness
cleaner bath
for press forming, the effect of pre-coat thickness on
press formability was studied using JIS Class-2 CP
titanium sheets having the properties shown in Table beyond this thickness, the improvement saturates.
2. Fig. 6 shows the results. Formability improves This result indicates that, within the range of
with increasing coating thickness up to 0.5μm, but deformation examined in this study, a coating as thin
as 0.5μm serves the purpose.
100 This lubricant pre-coating is applicable as-is to
PHEs; however, heat transfer via the coating
＋Wax
80 (Pre-coated)
deteriorates the heat transfer performance. To
Score (＝press formability) (％)

resolve this issue, the coating was modified so that

Resin＋Silica it can be removed by commonly-used alkali cleaning.
60 The removability of the coating was evaluated by
No coating
(Press Oil) the test described below.
40
Each lubrication coating was measured for its
coating quantity and subsequently immersed for a
predetermined time in a cleaning solution which is
20 generally recommended for titanium. After water
cleansing and drying, the coating quantity was
Compact size die JIS Class-2
measured. The degreasing solution was a weak
0
0.0 0.2 0.4 0.6 0.8 alkali, FINE CLEANER 4368 (manufactured by
Coefficient of dynamic friction (＝μk)
Nihon Parkerizing Co., Ltd.), which was diluted to a
Fig. 5 Effect of coefficient of dynamic friction on press concentration of 20g/L. The bath temperature was
formability
set at 60℃. The removal rate of the coating was
100 determined by Formula (6).
  Removal rate (%)＝100×(V0−V1)/V0 ……… (6)
wherein V0 (g/m2) and V1 (g/m2) are the coating
80
quantities before and after the cleaning, respectively.
Score (＝Press formability) (％)

Fig. 7 shows the relationship between the coating

60 removal rate and immersion time in the alkali
solution. The results indicate that coating removal
commences immediately after the coating is
40 immersed. The coating is removed completely
within 60 seconds. Since alkali cleaning is usually
20
conducted for 1 to 3 minutes, the coating has been
proven to be commercially applicable.
Compact size die JIS Class-2
0
0.0 0.5 1.0 1.5 2.0
3. Press formability of lubricant-pre-coated
Thickness of pre-coated layer (μm) titanium sheet for test production
Fig. 6 Effect of thickness of pre-coated layer on press
formability Sheets of CP titanium, having different tensile

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 22

 properties and pre-coated with lubricant, were tested Conclusions
for their press formability. The result was compared
with press oil lubrication and with film lubrication. Kobe Steel developed a lubricant-pre-coated
Fig. 8 compares the results. The large die was used titanium sheet to be used for PHEs, a major
for evaluating formability in a condition close to that application for commercially pure titanium. The pre-
of real press forming. The figure indicates that the coating can be applied to JIS Class-2 material that
pre-coated lubricant offers a favorable formability has a strength higher than that of the conventional
similar to that offered by lubricant film. It was also JIS Class-1, while retaining press formability close
found that the JIS Class-2 material, having strength to that achieved by press-oil lubrication. The
15% higher than that of JIS Class-1, exhibits a characteristics of the pre-coated titanium sheet have
formability similar to that of the JIS Class-1 material been introduced in this paper. This coating is also
that is press oil lubricated. designed to be environmentally friendly. The coated
layer, which is very thin, requires only light
100
degreasing, leaves much less residue and can be used
with much less concern than in the case of press
oil. The newly-developed technology is widely
Score (＝Press formability) (％)

80
applicable not only to PHEs, but also to various
JIS Class-1 press-formed products.
60
References
40
JIS Class-2
1) Kusamichi et al., Japan titanium industry and new technolog,
Large size die (1996), p.18. AGNE Gijutsu Center.
20 2) Processing Technology of Titanium, Japan Titanium Society
Press oil
Pre coated titanium (1992), pp. 82-86.
Polyethylene film 3) Takahashi et al., Development of High Lubricity Titanium
0
180 200 220 240 Sheet with Excellent Press Formability, Ti-2007 Science and
0.2％ Yield strength (MPa) Technology (2007), pp.1079-1082.
4) Moriguchi et al., R&D KOBE STEEL ENGINEERING
Fig. 8 Press formability of pre-coated JIS Class-1 and JIS REPORTS, Vol. 32, No. 1 (1982), pp.24-27.
Class-2, compared with using press oil and polyethylene 5) S. Ishiyama, Ph.D. thesis, Tohoku University, (1993). pp.13-
film as lubricant 78.

23 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

 Newly Developed Iron-based Powder Mixture, High-
density SEGLESS
, for High Density Compaction
Hironori SUZUKI＊1, Satoshi NISHIDA＊2, Takayasu FUJIURA＊3
＊1
Takasago Steel Powder Plant, Steel Powder Division, Iron & Steel Business
＊2
Steel Powder Sales Department, Steel Powder Division, Iron & Steel Business
＊3
Research & Development Planning Department, Technical Development Group

The newly-developed lubricant Kobelco Polyhydroxyl 100

Amide (KPA) consists of two different functional 98 2 press -

2 sinter
materials. Based on this feature, an iron powder 96
Powder
mixture with KPA lubricant has better powder flow at Reduce

Relative density
forging
94 lubricant
the powder stage and better ejection performance at
92 Warm compaction
the green stage. Since KPA lubricity is superior to that
of conventional lubricants, it is possible to reduce 90 Die wall lubricant
lubricant content without increasing the ejection 88
pressure. By utilizing this process, high-density powder
86 Conventional P/M
metallurgical parts can be made at a production cost
84
relatively lower than that of conventional methods. 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2
Relative production cost
Introduction Fig. 2 Production cost and typical density for various high
density processes
Iron-based sintered parts are produced by a
powder metallurgy process which comprises steps of engine performance. Among them, techniques such
mixing iron powder with additives, such as alloying as warm compaction and die-wall lubricated
powder and lubricant, compacting the mixture and compaction, which increase strength by decreasing
sintering the compact (Fig. 1). It is a low-cost high- porosity and thus densifying the sintered bodies, are
volume process for manufacturing complex-shaped effective because they are widely applicable to
products. Various efforts have been made to increase various sintered parts. These compaction techniques,
the strength of iron-based sintered parts, particularly however, require special ancillary equipment, such
for automotive engines, in response to the recent as heating and lubricant spray apparatuses. Thus,
need for reducing automotive weight and enhancing there has been a need for a more convenient method
of high density compaction.
Fig. 2 depicts the relationship between the
Powders Copper
Graphite
relative production costs and typical densities
Lubricant achieved by various powder metallurgical processes.
Conventional high-density processes such as warm
Iron powder Additives
compaction, die-wall lubricated compaction and
powder forging tend to cost more as the product
Mixing
density increases. Kobe Steel has found that a
functional lubricant can decrease the amount of
internal lubricant additive and achieve high density
Mixer
at relatively low cost without changing the existing
equipment. This paper introduces the features of the
Compaction, Ejection
functional lubricant, which exhibits excellent mold
releasability. Also introduced is an iron-based
powder mixture, containing the functional lubricant
and designed for high-density compaction.
Press

Dewax, Sintering 1. Role of lubricant in powder metallurgy

A lubricant for powder metallurgy is used for

Furnace reducing frictional forces among powder particles,
and between a compact and mold, particularly at the
P/M parts
time of the powder compaction and releasing of the
Fig. 1 Production process of powder metallurgy compact from the mold. Internal lubrication is a

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 24

 common approach for using lubricant, in which 20 Ejection pressure

Better lubricity
metal soap, such as zinc stearate (hereinafter referred High melt point
Compacting：686MPa

Ejection pressure (MPa)
metal soap
to as "Zn-st"), and/or wax resin, such as ethylene Size：φ25×15mm
bisstearamide (hereinafter referred to as "EBS") 15
are/is added in powder form to the base powder. Zn-st

Fig. 3 shows the relationship between lubricant

content and the theoretical maximum density 10
EBS
achieved by a commonly-used composition consisting Low melt point
Target fatty acid
of 2% copper powder and 0.8% graphite powder
with the balance of iron powder, and also the same 5
0 5 10 15 20 25 30 35 40
relationship found in the case of iron powder alone. Critical discharging diameter (mm)
This clearly shows that increasing the amount of Better flow
lubricant significantly decreases the density. It
Fig. 4 Relation between flowability and ejection pressure for
should be noted that the lubricant becomes useless various lubricant mixture 1)
once the compaction is completed and must be
removed from the powder compact. Thus, a smaller
amount of additive is preferable. To ensure lubricity, the other hand, remains solid during the powder
lubricant is typically added to iron powder in the mixing and exhibits excellent flowability; however,
amount of about 0.75 wt. percent. it does not accumulate sufficiently on the wall
Lubricant is known to significantly affect the surface and may exhibit poor mold releasability.
flowability of the mixture consisting of iron powder Thus, conventional lubricants have been unable to
and lubricant. It also affects the mold releasability, or satisfy both flowability and mold releasability at
the ejection force required to eject the compact of the same time.
the powder mixture from a mold after compaction. Targeting the development of a lubricant that
The flowability is evaluated by a unique Kobe can achieve both flowability and mold releasability,
Steel method and is represented by the minimum as shown by the arrow in Fig. 4, Kobe Steel has
bore diameter, known as the "critical discharging newly formulated a functional lubricant, Kobelco
diameter," which allows the discharge of subject Polyhydroxyl Amide (hereinafter referred to as
powder from a test container. After the comprehensive "KPA"), that selectively exerts the functions of
evaluation of a variety of commercially available assuring flowability during the powder mixing step
lubricants, flowability and mold releasability have and satisfying mold releasability at the time of
been found to be in a trade-off relation, as shown in compact ejection.
Fig.4 1).
A lubricant with a low melting point accumulates 2. Concept of functional lubricant, KPA
easily on the wall surfaces of a mold and facilitates
the release of compacts from the mold after KPA comprises two types of lubricants, both
compaction; however, it adheres strongly to iron polyhydroxylamides, but having different melting
powder during the powder mixing and may cause points2) (Fig. 5). A lubricant with a high melting
problems such as agglomeration and decreased point (no lower than 150℃), which contributes to
flowability. A lubricant with a high melting point, on flowability, is mixed with another lubricant with a
low melting point (no higher than 100℃), which
contributes to mold releasability, at an optimum
8.0 Fe-2％Cu-0.8％Gr ratio. These materials are also optimized for their
Fe qualities, such as particle size, to improve both
7.8 characteristics, flowability and mold releasability. It
Theoretical maximum
density (g/cm3)

should be noted that KPA is a wax-type lubricant

7.6
consisting only of organic materials, which decrease
7.4
the contamination of the sintering furnace and leave
less smearing on the surfaces of sintered bodies.
7.2 A mixture of iron powder and KPA was
compacted to evaluate dewaxing ability. The powder
7.0 compact was heated in a thermobalance to measure
0.0 0.2 0.4 0.6 0.8 1.0 1.2
Lubricant content (％) its weight loss during heating (Fig. 6). The result
Fig. 3 Effect of lubricant content on theoretical maximum indicates that the lubricant decomposes and gasifies
density as the temperature rises and is fully removed from

25 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

 of graphite.
Table 1 summarizes the typical properties of
these three powder mixtures, as well as those of
their green compacts. The apparent densities and
flow rates were measured according to JIS 2504 and
JIS 2502, respectively. Each powder mixture was
compacted into a cuboid (31.8mm×12.7mm×height
6.3mm) with a compacting pressure of 490MPa. The
green density was measured on each powder
compact. Also measured was its three-point bending
Amide 1 Amide 2 strength according to JIS Z2511. The mold releasability
Lower melting point Higher melting point
lubricant (＜373K) lubricant (＞423K)
was evaluated by the ejection force (peak value)
required to release each compact of the powder
mixture from the mold. The compacts were made
into solid cylinders (φ25×25mm, ea.) under three
different levels of compacting pressures; i.e., 490, 588
and 686MPa. Fig. 7 shows the results of the
releasability evaluations. The mixture containing
KPA requires a significantly smaller ejection force
compared with the conventional mixture and
Fig. 5 Conceptual diagram of functional lubricant KPA exhibits superior mold releasability. This effect is
enhanced at higher compacting pressures. The
Weight loss of lubricant KPA (％)

120
Temperature：Room temp. to 650℃
ejection force is decreased by about 20% at a
100 Heating speed：30℃/min compacting pressure of 686MPa.
Atmosphere：Nitrogen (100ml/min)
80
3.2 Lubrication mechanism of KPA
60

40 Several experiments were conducted to clarify

20 Steel powder＋0.6％KPA the lubrication mechanism of KPA, which exhibits
0
a superior mold releasability compared with Zn-st,
0 100 200 300 400 500 600 700 or with EBS, both of which are commercially
Heating temperature (℃) available. The viscoelasticity of the newly developed
Fig. 6 Weight loss of KPA lubricant during heating KPA and conventional EBS lubricants was measured

the powder compact in nitrogen atmosphere as the Table 1 Powder and green properties of KPA and other
temperature approaches 450℃. Almost no residue of lubricant mixture
lubricant was found after the dewaxing. Lubricant KPA Zn-st EBS
3
Apparent density (g/cm ) 3.27 3.49 3.24
3. Characteristics of KPA Flow rate (s/50g) 24.1 22.1 28.3

Critical flow diameter (mm) 22.5 22.5 30.0

3.1 Basic properties of powder mixture containing Green density (g/cm )*
3
6.92 6.90 6.91
KPA3), 4) Green strength (MPa)* 12.0 10.0 10.6
*Compacting pressure：490MPa
Three different mixtures were prepared, each
consisting of Kobe Steel's atomized iron powder, 300M＋2％Cu＋0.8％Gr＋0.75％Lubricant (SEGLESS)
12
300M (average particle size, 70μm), 2% of copper
Ejection pressure (MPa)

powder (average particle size about 30μm), and

11
0.8% of graphite powder (average particle size 20％
5μm), plus 0.75% of one of three lubricants, i.e., the reduction
10
newly developed KPA, Zn-st (average particle size
10μm) and EBS (average particle size 25μm). The 9 KPA
Zn-st
latter two lubricants, both commercially available EBS
8
and added in the same amount, were used as 400 500 600 700 800
reference materials. These compositions, called Compacting pressure (MPa)
"SEGLESS", were designed to prevent the segregation Fig. 7 Ejection pressure of KPA and conventional mixture

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 26

 at different temperatures, using an instrument for compacting pressure. This is considered to be the
measuring molten viscosity. The measuring reason for the more effective lubricity of KPA.
instrument comprises a pair of disks, rotating in Also studied were the behaviors of lubricants
opposite directions, and a layer of lubricant placed which migrate during compaction and ejection step
between them, the layer being maintained at a to accumulate on the surfaces of the powder compacts.
thickness of about 1mm. The viscoelasticity is Three powder mixtures, each containing iron powder
determined from the torque values obtained by and 0.75% of the three respective lubricants, were
rotating the disks at 62.5 rad/s (10Hz). Fig. 8 compacted under three different pressures. For each
compares the viscoelasticity of KPA to that of EBS. compact, the surface that was in contact with a punch
This figure indicates that the viscoelasticity of KPA during the compaction was observed. Fig.10 shows
gradually lessens as the temperature comes to exceed the results of energy dispersive X-ray (EDX) analyses
around 60℃, the temperature not exceeding the performed on the surfaces of the powder compacts.
melting point of KPA, and remains lower than that
of EBS. This suggest that, when ejecting the powder Load
compact from a mold, KPA, which has a lower Sliding
viscosity, can migrate more easily from the inside
of the powder compact and accumulate on the mold Thickness of lubricant
(10∼30μm)
wall surfaces to function better as a lubricant 5), 6).
Next, the static friction coefficients of KPA and 0.4
Sliding speed：100mm/min
EBS were measured using a sliding tester, HEIDON. Load：2∼3MPa
Each lubricant was applied on the surface of the steel
Coefficient of static friction
plate in a thickness of 10 to 30μm. Each sliding test 0.3 EBS
was conducted with a load of 2 to 3MPa at a sliding
speed of 100mm/min. The static friction coefficient
was determined from the peak value of the resistance
0.2
force during sliding for a specified distance. As KPA
shown in Fig. 9, the static friction coefficient of EBS
increases with the rise in the surrounding temperature,
while that of KPA appears to be independent of the 0.1
20 40 60 80 100
temperature. This difference becomes more significant Temperature (℃)
at temperatures above around 60℃. A static friction Fig. 9 Coefficient of static friction of KPA and conventional
coefficient represents the lubricity of the lubricant lubricant
itself. Even during compaction at room temperature, Die Steel powder
temperatures in some parts of the compact may
exceed 60℃ due to the friction heat caused by the

Lubricant

Lubricant
KPA Zn-st EBS

104 294
MPa

103
EBS
Viscoelasticity (Pa・s)

10 2 KPA
490
101 MPa

100

10−1 686
MPa

20 40 60 80 100 120 140 160 180

Temperature (℃)
Fig.10 Lubricant distribution on surface of green compact by
Fig. 8 Viscoelasticity of KPA and conventional lubricant EDX

27 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

The yellow portions are where carbon, contained in higher by 0.2g/cm3 than that achieved by the
each lubricant, is detected. These results indicate conventional SEGLESS. Fig.12 shows the forces
that, for all of the compacting pressures, larger required to eject the powder compacts of the various
quantities of KPA accumulate on the compact surfaces mixtures. In both the cases, the ejection force
than in the case with the other two lubricants. increases with increasing compact density. When
It has been concluded from the above experiments comparing the two mixtures at a given compact
that KPA not only has a low static friction coefficient, density, High-density SEGLESS with reduced
which improves its lubricity, but also has a low lubricant exhibits an ejection force comparable to that
viscosity that enables the lubricant to effectively seep of conventional SEGLESS, proof that the amount
to the compact surfaces. The resulting synergistic of lubricant additive can be decreased without
effect provides KPA with a mold releasability that is sacrificing ejection force.
superior to that of other lubricants. Finally, both of the powder mixtures were
compacted under different pressures into a ring
3.3 Characteristics of High-density SEGLESS shape with an outer diameter of 64mm, inner
containing KPA diameter of 24mm and height of 10mm. The ring-
shaped compacts were sintered at 1,120℃ for 30
KPA, with its excellent mold releasability, is minutes in a nitrogen atmosphere containing 5%
expected to reduce the burden imposed on molds. of hydrogen to produce sintered bodies. The
This will extend the life of the molds and enable characteristics of the sintered bodies were compared.
it to be applied to parts having complicated shapes. Compressional load was applied by a press in the
One way of using KPA is to decrease the amount radial direction of each ring-shaped sample. The
added to a level that attains the ejection force radial crushing strength was determined from the
achieved by conventional lubricants, while increasing load that caused fracturing in each ring. The results
the density of compacts. The high-density process are shown in Fig.13. In the case of High-density
with a reduced amount of lubricant additive does not SEGLESS, the strength of the sintered body increases
call for additional equipment and, thus, can be
established with relative ease. Such a process emits Table 2 Powder properties of high density SEGLESS and
much less gas during dewaxing and is more conventional SEGLESS
environmentally friendly. This section introduces the High density SEGLESS Conventional SEGLESS
features of a newly developed powder mix, "High- Apparent density
3.49 3.46
density SEGLESS", which contains KPA and halves (g/cm )
3

the amount of lubricant additive compared with Flow rate

23.1 26.5
conventional technology. (s/50g)

Two types of powder mixtures were prepared:

i.e., High-density SEGLESS, consisting of Kobe Steel's 7.35
Green density (g/cm3)

7.30
high-compressibility iron powder, 300NH, 2% of
7.25
copper powder (average particle size, about 30μm), 7.20
0.2g/cm3
0.8% of graphite powder (average particle size, 5μm) 7.15
and 0.4% of KPA; and a conventional SEGLESS, 7.10
containing 0.8% of conventional lubricant, EBS 7.05 High density SEGLESS
7.00 Conventional SEGLESS
(average particle size, 25μm). Table 2 summarizes 6.95
the powder properties of High-density SEGLESS 400 500 600 700 800
Compacting pressure (MPa)
and the conventional SEGLESS. Both have a similar
apparent density, however, High-density SEGLESS Fig.11 Compressibility curve of high density SEGLESS
has a smaller amount of lubricant added and thus has
18
a flow rate improved by 3s/50g compared with that High density SEGLESS
Ejection pressure (MPa)

16
of the conventional SEGLESS. 14
Fig.11 compares High-density SEGLESS with the 12
conventional SEGLESS, in terms of the relationship 10
8 Compacting pressure
between the compacting pressure, required to compact 6 Conventional SEGLESS 490MPa
the mixture into a solid cylinder of φ25×25mm, and 4 588MPa
the resulting compact (green) density. High-density 2 686MPa

SEGLESS exhibits an improved compact density, 0

6.9 7.0 7.1 7.2 7.3 7.4
7.30g/cm3 (compacting pressure, 686MPa), which is Green density (g/cm3)
a result of the reduced amount of lubricant and is Fig.12 Ejection pressure of high density SEGLESS

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 28

 Radial crushing strength (MPa)
1,200 metallurgy, and the features of the powder mixture
1,100 High density SEGLESS comprising iron powder and the lubricant:
1,000 16％ 1) KPA, consisting of two types of lubricants with
900 Compacting pressure different melting points, can improve both
800 Conventional SEGLESS 490MPa flowability and mold releasability at the same
588MPa
700 686MPa
time. Such an improvement has conventionally
600
been difficult to achieve.
6.9 7.0 7.1 7.2 7.3 2) The iron-based mixed powder containing KPA
Sintered density (g/cm3)
exhibits an excellent mold releasability, which
Fig.13 Radial crushing strength of sintered high density
reduces the burden imposed on molds and is
SEGLESS
suitable for use in parts having complicated, thin-
100 walled shapes.
Apparent hardness (HRB)

95 3) High-density SEGLESS, containing an amount of

90 High density SEGLESS KPA decreased to a level that exhibits the same
85 7％
ejection force as exhibited by conventional
80 Compacting pressure SEGLESS, enables the production of higher
75
Conventional SEGLESS 490MPa density parts without additional equipment. This
588MPa
70 686MPa
responds to the needs for parts with higher
6.9 7.0 7.1 7.2 7.3 strength and lower weight, features receiving
Sintered density (g/cm3) much attention in recent years.
Fig.14 Apparent hardness of sintered high density SEGLESS 4) The newly developed lubricant is applicable to a
wide variety of products, such as pure-iron based
with increasing density, and, for a given compacting machine parts, high-strength parts made from
pressure, it increases by about 16% compared with alloy powder with inferior compactability and
conventional SEGLESS. Fig.14 shows the apparent magnetic iron powder requiring higher density.
hardness, measured on the Rockwell Hardness B-
Scale, of each sintered body. The hardness was References
determined from the average of the values measured
at three points on the upper surface and another 1) H. Suzuki et al., R&D KOBE STEEL ENGINEERING
three points on the lower surface. The hardness also REPORTS, Vol.55, No.3 (2005), pp.64-67.
2) S. Nishida et al., Proceeding of 2006 Powder Metallurgy World
increases with increasing density in the same manner Congress (2006), PartⅡ, pp.731-732.
as in the case of strength; at a compaction pressure 3) H. Suzuki et al., Japan Society of Powder and Powder Metallurgy,
of 686MPa, it increased by about 7% as compared Abstracts, Annual Spring Meeting (2005), p.124.
with conventional SEGLESS. 4) H. Suzuki et al., Proceeding of the 2007 International Conference
on Powder Metallurgy & Particulate Materials (2007), Part 3,
pp.11-18.
Conclusions 5) Y. Ito et al., Japan Society of Powder and Powder Metallurgy,
Abstracts, Annual Autumn Meeting (2007), p.81.
The following summarizes the characteristics of a 6) Y. Ito et al., Proceeding of the 2008 World Congress on Powder
newly developed functional lubricant, Kobelco Metallurgy & Particulate Materials (2008), Part 3, pp.41-49, 74.
Polyhydroxyl Amide (KPA), developed for powder

29 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

 Newly Developed Iron Powder for Highly Efficient Dust
Cores
Hirofumi HOJO＊1, Nobuaki AKAGI＊1, Tetsuya SAWAYAMA＊2, Hiroyuki MITANI＊3
＊1
Takasago Steel Powder Plant, Steel Powder Division, Iron & Steel Business
＊2
Steel Powder Division, Iron & Steel Business
＊3
Materials Research Laboratory, Technical Development Group

Dust cores are made by compacting insulated magnetic than by eddy current loss. This makes it difficult to
powder. One of the advantages of dust cores is magnetic take advantage of the reduced eddy current loss of
isotropy, which enables three-dimensional magnetic dust cores, hence the desire for a dust core with
circuit designs. Such designs are expected to reduce the decreased hysteresis loss.
size and weight of magnetic parts. There is a strong The present paper describes technologies, including
demand to reduce core loss so that the cores can be used the determination of optimum temperature for
more for magnetic parts, such as motor cores. This magnetic annealing to relieve strain for reducing the
paper reports on methods for reducing iron loss in dust iron loss of dust cores. This paper also includes the
cores and on the magnetic properties of the dust cores development of a highly heat-resistant insulation
consisting of improved powder made by a new method. coating for the optimum magnetic annealing, as
A dust core with an iron loss of 28.6W/kg was produced. well as the effect of powder particles and crystal
This core is comparable with ones made of laminated grain sizes on iron loss. Also introduced are the
steel sheets. characteristics of a powder compact with reduced
iron loss, the compact having been developed on the
Introduction basis of the knowledge that was obtained.

To decrease the environmental burden and to save 1. Factors determining iron loss of dust core
energy, there are growing needs for electromagnetic
components with reduced loss. Components such as In general, iron loss consists of eddy current loss
photovoltaic power conditioners and the reactors and hysteresis loss. The eddy current loss is further
that are used for boost converters of hybrid electric classified into intra-particle eddy current loss, caused
vehicles (HEVs) and electric vehicles (EVs) require by the eddy current flowing inside each individual
magnetic materials that have low magnetic loss and particle, and inter-particle eddy current loss, caused
can be used at higher frequencies. In addition, the by the eddy current flowing across the particles
electric motors used for HEVs and EVs call not only throughout the component.
for reduced magnetic loss, but also for smaller size The hysteresis loss of a material, on the other
and lighter weight. hand, corresponds to the looped area in the BH
Dust cores, made by compacting insulated soft- (magnetic flux density vs. magnetic field) curve, in
magnetic powder, have the advantage that they can which the size of the loop correlates strongly with the
suppress eddy current loss because the constituent material's coercive force (i.e., coercivity). Thus,
particles are electrically insulated. The eddy current decreasing the coercivity decreases the hysteresis
loss is proportional to the second power of frequency. loss. The factors determining coercivity are particle
Thus, components used at higher frequencies can composition (solid solution elements), precipitates,
better exploit this advantage. Because of this, dust inclusions, crystal grain boundaries, powder surfaces
cores are progressively being used for applications and dislocations (strains), i.e., the factors that inhibit
such as the reactors for the boost converters of HEVs. the movement of domain walls.
In addition, dust cores have isotropic magnetic The factors determining the intra-particle eddy
characteristics. This enables the designing of new current loss include particle composition (solid
magnetic circuits with three-dimensional structures, solution elements), which affects the electrical
which have been difficult to design using conventional resistivity of each particle, and particle size, which
magnetic cores made of laminated steel sheets defines the area within which the eddy current flows.
having magnetic anisotropy. Because of this, various The factors determining the inter-particle eddy
efforts are being made to exploit dust cores for the current loss include the electrical resistivity of the
downsizing and weight reduction of components, insulation coating, which affects the inter-particle
including electric motors1), 2). Such components, insulation (Fig. 1).
however, are typically driven by the commercial Among these factors, the coercivity is decreased
frequency, which is so low that the iron loss is more significantly as the temperature for magnetic
governed predominantly by hysteresis loss, rather annealing, performed to relieve strain, rises.

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 30

 impurity properties were measured on these samples, using a
iron
powder particle BH curve tracer (BHS-40S by Riken Denshi Co., Ltd).
surface Furthermore, the sample cross-section was etched
insulate
layer with NITAL (an etching solution consisting of ethyl
strain alcohol and nitric acid) so as to be subject to
(dislocation) grain boundary inter-particle intra-particle
eddy current eddy current metallographic observation under an optical
control factor of control factor of microscope.
hysteresis loss eddy current loss

Fig. 1 Control factor of iron loss 2.2 Results and discussions

Conventional dust cores, however, suffer from the Fig. 2 shows the relationship between the
problem of the lack of heat resistance of the temperatures for magnetic annealing and the
insulation coating on the particle surfaces, which resulting coercivities of the samples. The coercivity
causes the fracture of inter-particle insulation during decreases as the annealing temperature approaches
annealing at elevated temperatures, resulting in 973K. This is considered to be caused by the relief of
increased inter-particle eddy current loss. Another strain in the compacts by the magnetic annealing.
known technique for decreasing coercivity is to The sample annealed at 1,073K exhibits a coercivity
increase particle size. Intra-particle eddy current loss, higher than that of the sample annealed at 973K.
however, is proportional to the second power of There is only a slight difference in coercivity between
particle size; hence particle size is a factor controlling the compact annealed at 873K and the one annealed
trade-off characteristics. at 973K. This result indicates that the coercivity-
From the above, it can be seen that an important decreasing effect of annealing becomes nearly
issue in reducing the iron loss of a dust core is to saturated at about 873K.
suppress its eddy current loss and at the same time Fig. 3 are optical micrographs of the cross-sections
to decrease its hysteresis loss. of two compacts, i.e., one before magnetic annealing,
and the other after magnetic annealing at 1,073K. The
2. Optimum annealing temperature for relieving crystal grains of the compact annealed at 1,073K are
stress from compacts finer than they were before magnetic annealing. This
shows recrystallization. In general, the smaller the
Dust cores are compacted in dies. Therefore, grain size, the higher the coercivity. Thus, the
the powder particles are plastically deformed and increase in coercivity of the compact annealed at
have a large amount of strain introduced. Strain 1,073K is attributable to the grain refining caused by
(dislocation) is a factor which increases coercivity recrystallization. If we examine this in more detail, in
and can be relieved by heat treatment. A study was the temperature range no lower than 873K, the
conducted to determine the optimum annealing
temperature for strain relief. 350
Coersivity (A・m−1)

2.1 Experimental method 300

250
Pure iron powder (300NH, manufactured by
200
Kobe Steel; average particle size, approximately
100μm) is mixed with 4.0mass% of highly-pure 150
alumina fine powder (manufactured by Showa
100
Denko K.K.; average particle size, approximately 0 200 400 600 800 1,000 1,200
Heat treatment temperature (K)
1.3μm), along with an organic binder. The particles
of the highly-pure alumina powder are adhered to the Fig. 2 Relationship between coersivity and annealing
temperature
surface of iron powder particles, forming insulation
layers. The powder mixture was compacted into ring-
shaped samples, each having an outer diameter of
45mm, inner diameter of 33mm and height of 5mm,
by die-wall lubricated compaction with a compacting
pressure of 980MPa. Zinc stearate was used as the die
wall lubricant. Each compact was annealed in a
nitrogen atmosphere at a temperature in the range of Fig. 3 Cross sectional microstructure of before annealed
673 to 1,073K for 7.2ks. The direct-current magnetic (left) and annealed at 1,073K(right) compact

31 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

coercivity-decreasing effect of magnetic annealing 20

Eddy current loss, We1.0T/400Hz(W・kg−1)

lessens, and the upper limit annealing temperature,
effective in decreasing coercivity, is regarded as to be 15

approximately 973K. In addition, magnetic annealing

10
at 1,073K has turned out to adversely affect the
coercivity reduction.
5
From the above, the optimum temperature for
magnetic annealing of dust core made from pure iron
0
powder is concluded to exist in the temperature 0 100 200 300 400

range from 873K to 973K. Electric resistivity (μΩ・m)

Fig. 4 Relationship of eddy current loss and electric resistivity

3. Heat resistance of insulation coating
1,000
The magnetic annealing at the optimum temperature

Electric resistivity (μΩ・m)

determined in section 2 is effective in decreasing 100
coercivity and thus reducing hysteresis loss;
however, conventional insulation coatings suffer 10
from the problem of their heat resistance not being coating-A
high enough. The following outlines a heat resistant 1 coating-B
coating developed to solve this problem. coating-C

0
200 300 400 500 600 700 800 900 1,000
3.1 Experimental method Heat treat temperature (K)
Fig. 5 Relationship between electric resistivity and annealing
Three powder samples mainly consisting of pure temperature
iron powder, 300NH, were prepared. Each powder
sample was coated with two layers, i.e., one resistivities on the annealing temperature of powder
inorganic layer and one organic layer. Each inorganic compacts, each coated with two layers, one layer
layer was formed from one of three phosphoric acid- consisting of one of the three inorganic materials
based solutions (A, B or C) for insulation. The organic made from phosphoric acid solutions and the other
layer consists of silicone resin having heat resistance. layer consisting of silicone resin. All the coatings
The coated powder samples were compacted by die- exhibit a decrease in electrical resistivity with rising
wall lubricated compaction with a pressure of annealing temperature. The decrease is considered to
980MPa into plate-shaped specimens, each having be due to the deterioration of the coatings. The two-
dimensions of 12.7mm×31.8mm×5mm. The layered coating, consisting of a layer of inorganic
compact specimens were annealed in a nitrogen material (A) and a layer of silicone resin, exhibits a
atmosphere at a temperature in the range of 773 to smaller decrease in electrical resistivity and has an
873K for 1.8ks. Their electrical resistivities were electrical resistivity higher than 100μΩ・m at an
measured by a four-terminal method. annealing temperature of 873K, the resistivity required
at the optimum magnetic annealing temperature.
3.2 Results and discussion There are reports on the improvement in heat
resistance of inorganic coating made from a
The electrical resistivity of an insulation coating phosphoric acid-based solution. Such reports include
varies depending on the coating material, coating one focusing on the vitrification temperature of
thickness and magnetic annealing temperature. coating3) and another focusing on the diffusion of
These factors, determining the electrical resistivity elements4). However, the cause of the high heat
of insulation coating, were changed to study the resistance of the two-layer coating, consisting of the
eddy current losses of samples with various electrical inorganic layer (A) and silicone resin, as reported in
resistivities (Fig. 4). The result indicates that an this paper, requires further investigation.
electrical resistivity higher than about 100μΩ・m Conventional insulation coatings have low heat
effectively suppresses the inter-particle eddy current resistance, limiting the magnetic annealing
loss. It is thus concluded that the coating used for a temperature to around 773K. The newly developed
dust core must have an electrical resistivity that two-layered coating allows magnetic annealing at
causes the overall resistivity of the compact to be 873K, a temperature high enough for the strain relief
higher than about 100μΩ・m. of dust cores.
Fig. 5 compares the dependence of electrical

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 32

 4. Effect of powder particle size and crystal grain distribution width of the powder samples were A) 0
size on iron loss ∼250μm; B) 45∼250μm; C) 63∼250μm; D) 75∼250
μm; E) 106∼250μm; F) 100∼250μm; and G) 180∼
Studies were conducted on the effects of powder 250μm. Fig. 8 shows the coercivities of the compacts
particle size and crystal grain size on iron loss. The made from these powders. It has turned out that
powder particle size, in particular, affects both the coercivity is effectively decreased by growing crystal
hysteresis loss and eddy current loss. Thus, a study grain and removing finer particles from the powder
was conducted from the two aspects of hysteresis to increase the average particle size.
loss and eddy current loss. Also outlined is the If all the powder particles become single crystals
relation between the powder particle size and crystal as a result of the powder heat treatment for crystal
grain size, which are inseparably connected. grain growth, no further increase in the average grain
size is expected. Thus, finer particles serve to
4.1 Experimental method decrease coercivity. When removing finer particles
by sieving, it is effective to remove particles smaller
Pure iron powder, 300NH, was heat-treated in than the crystal grain size; however, fine crystals in
powder form. Each heat-treatment cycle was coarse particles still inhibit the decrease in coercivity.
conducted in a nitrogen atmosphere at 1,243K for Thus, in order to decrease coercivity, it is
5.4ks and the cycle was repeated three times to grow important to combine the removal of fine particles
crystal grains. The resulting powder was sifted using (coarsening) and the growth of crystal grains.
sieves with mesh sizes of 180, 150, 100, 75, 63 and It should be noted that increasing the particle size
45μm. Powder samples were prepared with the increases the intra-particle eddy current. Therefore,
finer particles removed sequentially, as described in particle size should be selected by incorporating the
4.2. Each powder sample thus obtained was
compacted into a ring-shaped specimen having a
compact density of 7.60Mg/m3 by die-wall lubricated
compaction. Each compact specimen was annealed
in a nitrogen atmosphere under one of two
conditions, i.e., 823K for 1.8ks, or 873K for 1.8ks. The
annealed specimens were measured for their direct
current magnetic properties. Iron loss was measured Fig. 6 Cross sectional microstructure of powder with(right)
by an automatic magnetic measurement apparatus and without(left) grain growth treatment
(manufactured by Yokogawa Electric Corporation).
260
The conditions for iron loss measurement were the
Coersive force (A・m−1)

240
excitation magnetic flux density, 1.5T, and excitation 220
frequency, 200Hz. 200
180
160
4.2 Results and discussion 140
120
In a magnetic material, interfaces such as grain 100
annealed at annealed at annealed at annealed at
boundaries and powder surfaces provide sites for 823K 873K 823K 873K
preventing domain wall migration. To study this normal crystal grain powder large crystal grain powder

behavior, pure iron powder was heat treated to Fig. 7 Coersive force comparison of crystal grain size
grow its crystal grains. Fig. 6 shows the cross-
sectional micrographs of the powders before and 300

after the heat treatment. The heat treatment has annealed at 823K
Coercive force (A・m−1)

been shown to grow the crystal grains. 250 annealed at 873K

Fig. 7 compares the coercivities of the compacts of

200
the powders before and after the heat treatment
applied for crystal grain growth. The figure indicates
150
that coercivity is decreased as a result of the grain
growth.
100
After the grain growth heat treatment, the powder A B C D E F G
0∼ 63∼ 106∼ 180∼
was sifted using sieves with mesh sizes of 180, 150, 45∼ 75∼ 150∼
100, 75, 63, 45μm. Powder samples were prepared Particle size (μm)
with finer particles removed sequentially. The size Fig. 8 Relationship of particle size and coersive force

33 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

 45
(a)
To investigate the characteristics of compacts
annealed at 823K
made from the newly developed powder, pure iron
Core loss, W1.0T/400Hz annealed at 873K
(W・kg−1) 40 powder (i.e., starting material, 300NH), was classified
using a sieve with 150μm mesh openings to remove
fine particles. The remaining coarse powder was
35
treated to grow its crystal grains. The treated powder
was coated by two-layer insulation, consisting of a
30 phosphoric inorganic layer (A) and a silicone resin
A B C D E F G
0∼ 63∼ 106∼ 180∼ layer. This powder was heated to 403K, compacted
45∼ 75∼ 150∼
Particle size (μm)
by die-wall lubricated compaction at a pressure of
1,176MPa, and was annealed in a nitrogen atmosphere
45
Eddy current loss (W・kg−1)

(b) hysteresis loss annealed at 823K at 873K for 1.8ks to prepare the compact specimen.
Hysteresis loss (W・kg−1),

40
(hollow mark)
35 The magnetic properties and iron losses of the
30 specimens are shown in Table 1 and Fig.10,
25
annealed at 873K (solid mark) respectively. Also shown for comparison are the
20
15 characteristics of a conventional material and
eddy current loss
10 electrical steel sheets. The ring-shaped specimens for
5 the electrical steels were prepared by wire cut. The
0
A B C D E F G conventional dust core in Fig.10 was made from
0∼ 63∼ 106∼ 180∼
45∼ 75∼ 150∼ 300NH with conventional insulation coating applied,
Particle size (μm) after which it was annealed at 773K for 1.8ks.
Fig. 9 Relationship between particle size and core loss(a), The newly developed powder exhibits significantly
and between particle size and hysteresis, and eddy reduced iron loss compared with the conventional
current loss(b)
materials. It shows reduced iron loss in the
frequency range above 200 to 300Hz, compared
decreasing effect of hysteresis loss and increasing with a conventional electrical steel sheet, JIS 50A400.
effect of eddy current loss. It also shows reduced iron loss in the frequency
Fig. 9 (a) represents the iron losses of compacts range above 700∼1kHz, compared with JIS 35A360.
made from iron powders with or without fine Thus the newly developed powder is expected to be
particles, and Fig. 9 (b) represents the respective iron used for applications such as motors that rotate at
loss, separated into hysteresis loss and eddy current relatively high speed.
loss. As shown in Fig. 9 (b), removing fine particles
is effective in decreasing both coercivity and
hysteresis loss and has a minimal effect in increasing 100 developed
dust core
eddy current loss within the particle size distribution conventional
investigated. Therefore, the iron loss shown in Fig. 9 dust core
Core loss (W・kg−1)

electrical sheet
(a) monotonically decreases as a result of removing JIS 50A400
electrical sheet
fine particles. 10 JIS 35A360

5. Characteristics of newly developed iron powder

with reduced iron loss

The following introduces the iron powder with 1

10 100 1,000
reduced iron loss developed on the basis of the Frequency (Hz)
knowledge obtained. Fig.10 Comparison of core loss

Table 1 Properties of developed material

size heat flux
core loss
transverse
maximum
distribution density treatment density coersivity
(1.0T/400Hz)
rupture
permeability
(sieve size) temperature (@8,000A/m) strength
3
(μm) (Mg/m ) (K) (T) (A/m) (W/kg) (MPa)
developed 150∼250 7.69 873 1.66 669 134 28.6 76
dust core
conventional 0∼250 7.61 823 1.58 446 243 50.4 82

electrical JIS 50A400 − − as cutted 1.8 6,213 60 35.5 −

sheet JIS 35A360 − − as cutted 1.79 6,532 59 22.9 −

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 34

 Conclusion crystal grains and powder particles, rather than
controlling them independently.
As a means for decreasing coercivity and ・Combining the newly-developed heat-resistant
hysteresis loss, a new iron powder was developed insulation coating with crystal grain coarsening
focusing on magnetic annealing, powder particle size and particle coarsening (removal of fine particles)
and crystal grain size. The following was found: enable the production of a dust core having iron
・The optimum magnetic annealing temperature loss comparable with that obtained by conventional
for dust cores made from pure iron powder lies electrical steel sheets. Dust cores have the feature
in the range from 873 to 973K. Annealing at of being able to suppress eddy current loss at high
temperatures above this range adversely increases frequencies. Thus, the newly developed powder
coercivity. This is partially due to the refinement of can be used as a material with iron loss lower than
crystal grains caused by recrystallization. that achieved by electrical steel sheet in the high-
・An electrical resistivity higher than about frequency range (several hundred Hz to 1kHz or
100μΩ・m is required to suppress the magnetic above).
loss caused by inter-particle eddy current. An
insulation coating was developed which maintains References
the required resistivity even after magnetic
annealing at 873K. This coating has made it 1) Lars Hultman et al., Advances in Powder Metallurgy &
possible to decrease coercivity and, therefore, Particulate Materials, (2002), pp.14-26.
2) Y. Enomoto et al., The transactions of the Institute of Electrical
hysteresis loss. Engineers of Japan, D, Vol.129, (2009), pp.1004-1010.
・The effects of crystal grain size and powder 3) S. Tajima et al., Journal of the Japan Society of Powder and
particle size on the coercivities of compacts have Powder Metallurgy, Vol.53, (2006), pp.290-296.
been clarified. A significant reduction of coercivity 4) S. Tajima et al., Journal of the Japan Society of Powder and
is made possible by combining the size effects of Powder Metallurgy, Vol.52, (2005), pp.164-170.

35 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

 Overseas Deployment of Wire Rod & Bar Manufacturing
and Wire Rod Secondary Processing
Goro AKAISHI＊1, Noriaki HIRAGA＊2
＊1
Wire Rod & Bar Products Marketing & Technical Service Department, Iron & Steel Business (currently with Wire Rod Production Department,
Kakogawa Works)
＊2
Wire Rod & Bar Products Marketing & Technical Service Department, Iron & Steel Business

The demand for special steel wire rods and bars has increasingly important to ensure a stable supply of
been increasing as Japanese auto makers continually materials of the same quality as those delivered in
expand their overseas operations, increasing the local Japan, using a similar delivery system.
procurement of automotive parts. Kobe Steel now has This paper reports on Kobe Steel's overseas
a technical tie-up with two special steel mills in Europe deployment of sites for fabricating wire rods and
and North America so that high quality wire rod and bars made of special steel. Also described is the
bar can be supplied to auto makers and parts makers. deployment of overseas sites for secondary processing
Kobe Steel has also constructed a system to provide for products.
the secondary processing of wire rod products for
Japanese parts makers, mainly by establishing secondary 1. Kobe Steel's approach toward establishing
processing companies in Asia and North America. overseas production sites

Introduction Special steel wire rods and bars have a wide

variety of applications. In general, they are used not
The automobile trade dispute between Japan and only for parts such as bearings, springs and gears,
US caused Japanese automakers to accelerate their which are subject to continuous dynamic load, but
local auto production in North America, a major also for applications such as structural steel, which
consumption area. This has led Japanese automakers are used under static load but are exposed to severe
to expand their manufacturing bases in other areas, cold work during the manufacturing process. In
such as Europe and Asia, where the demand for other words, the wire rods and bars must have high
automobiles has been increasing. Some regions may quality surfaces and interiors in response to the
have significance as operational bases for exporting performance requirements for end products and
because of low manufacturing costs. In 2009, the manufacturing processes.
overseas production of Japanese automakers As described, many Japanese auto parts employ
exceeded their domestic production1) (Fig. 1). Many wire rods and bars made of customized special
Japanese parts manufacturers have moved into steels. Thus, the required qualities vary item by item.
North America, Southeast Asia and China and are To meet these quality requirements, detailed and
expanding their on-site production against the flexible management is needed; and to that end, the
backdrop of the rising yen. overseas production sites must have manufacturing
Without a doubt, Japanese auto parts are and quality control, production planning and technical
supported by customized steel of consistent quality, services comparable to what has been implemented
as well as secondary processing products. For the in Japan. Kobe Steel proceeds with this implementation
overseas manufacturing bases, it has thus become while keeping the following aspects in mind.

1.1 Business form

14,000
Japan
Number of automobile production

12,000 Overseas Even if made of special steel, steel bars are hot-
10,000
forged as-is in many cases. In contrast, wire rods are
(×1,000 units)

cold-forged plastically into more near-net shapes,

8,000
requiring a secondary process that gives the wire
6,000 rods homogeneous structures and sizes suitable for
cold forging. This make the role of secondary
4,000
processing increasingly important.
2,000 Kobe Steel's overseas deployment depends on
0 the type of product. The steel bars are made by
2005 2006 2007 2008 2009
Year
local manufacturers of special steel with technical
Fig. 1 Domestic and overseas automobile production by assistance agreements. On the other hand, secondary
Japanese auto makers processing sites are established for wire rods. The

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 36

 wire rods are either imported from Kobe Steel in companies in close relationships with Kobe Steel,
Japan, or procured from special steel manufacturers have jointly established firms to provide bases for
who are located near the secondary processing sites secondary processing. These bases are provided with
and are under technical assistance agreements. state-of-the-art facilities, as well as with engineers
and managers who are stationed locally to transfer
1.2 Manufacturing & quality control and the technologies for production/quality control and
production management to implement management know-how.

The impetus behind the need for item-by-item 1.3 Technical service
quality control is the downsizing and weight
reduction of Japanese automobile parts, which It is extremely important to listen to customers'
require materials that have high strength and can voices to understand how our products are being
withstand heavy working. In particular, parts that used and to determine if the customers are satisfied,
are subject to heavy cold working call for materials or if they have any problems. Visual confirmation of
without defects. Thus, the local industry demands the usability of the products is also important.
steel with no surface defects and no non-metallic Determining how our products should be made
inclusions near the surface, which are harmful to the through such continuous dialogue with customers is
quality. Also in demand is secondary processing that the most effective way of strengthening product
optimizes the microstructure for cold working. To competitiveness.
meet these requirements, it is necessary to fully Kobe Steel deploys capable human resources to
understand what is going on during the production the bases and resident offices in charge of the various
processes and to implement detailed quality control; regions, where they collect relevant information so as
however, these are technologies supported by to ensure smooth communication with customers,
accumulated data and experience and are impossible thus adding value to the products.
to learn in a short time.
Thus, Kobe Steel focuses on selecting local 2. Deployment of manufacturing bases by regions
special steel manufacturers with high technological
competence, especially in quality control. Kobe Steel Kobe Steel currently has manufacturing bases at
also takes the approach of stationing technical teams important locations in North America, Europe and
at special steel manufacturers equipped with first- Asia (Fig. 2) to respond to customers' needs.
class facilities. The following describes the market trend and
Furthermore, Kobe Steel and secondary processing status of deployment in each region.

● ASCOMETAL
● Republic Engineering Products

★ Jiangyin Sugita Fasten

Spring Wire Co., Ltd.

★ Kobe Special Steel Wire

Products (Pinghu) Co., Ltd. ● Kobe Steel, Ltd.

★ Kobe Wire Products

(Foshan) Co., Ltd.

★ Grand Blanc Processing L.L.C.

★ Mahajak Kyodo Co., Ltd.

★ Kobe CH Wire
(Thailand) Co., Ltd.
●：Special steel company
★：Secondary processing company

Fig. 2 Location of overseas special steel mill which have a technical tie-up with Kobe Steel and overseas secondary processing
companies

37 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

2.1 North America Republic Engineered Products (REP), Ohio, USA, for
its production of special steel wire rods and bars and
The automobile industry in North America was has licensed it to produce wire rods for suspension
hit hard by the Lehman Shock in 2008. Two of the Big springs developed by Kobe Steel. Kobe Steel also has
Three companies filed for protection under federal a base at Grand Blanc Processing L.L.C. (GBP),
bankruptcy laws. In this environment, Japanese Michigan, USA, to produce steel wire for cold
automakers have increased their market share, heading and for bearing parts such as balls, rollers
thanks to their cars featuring high fuel economy and and needles. Currently, both the steel products and
requiring minimum maintenance1) (Fig. 3). secondary processing products are supplied not only
Kobe Steel provides technical assistance to to Japanese customers, but also to local consumers.
The company outlines of REP and GBP are included
25,000 40.0 in Table 1 and 2, respectively.
Other than Japanese auto
Number of automobile production

Japanese auto 35.0

Ratio of Japanese auto
20,000 2.2 Europe

Ratio of Japanese auto (％)

30.0
(×1,000 units)

25.0
15,000 Many steel manufacturers in Europe exceed or
20.0 even equal Japanese manufacturers in technological
10,000 competence and management know-how. They are
15.0

10.0
capable of independently producing special steel
5,000 wire rods and bars with high quality, once they learn
5.0
the Japanese method of quality control.
0 0.0 Kobe Steel and ASCOMETAL agreed on technical
2005 2006 2007 2008 2009
Year cooperation including a license exchange for
Fig. 3 Automobile production in North America producing steels developed by their respective

Table 1 Overseas technical tie-up mills and their outline

Company name Republic Engineered Products ASCOMETAL
Head office Lorain, OH, USA Paris, France
Tie-up year 1999 2002
Main factory Lorain Canton Lackawanna Hagondange Fos sur mer Le Cheylas
Blast furnace Electric furnace Bar mill Electric furnace Electric furnace Electric furnace
Basic oxygen furnace LF Bar inspection LF LF LF
LF Vacuum degassing Vacuum degassing Vacuum degassing Billet CC
Vacuum degassing Bloom CC Bloom CC Ingot casting Bar mill
Main facilities Bloom CC Billet CC Bar mill Billet inspection Bar inspection
Billet CC Bar inspection Wire rod mill
Billet inspection
Bar mill
Wire rod mill

Table 2 Overseas wire rod secondary processing companies and their outline
Grand Blanc Processing Kobe CH Wire (Thailand) Kobe Wire Products (Foshan) Jiangyin Sugita Fasten Spring Kobe Special Steel Wire
Mahajak Kyodo Co., Ltd.
Company Name L.L.C. Co., Ltd. Co., Ltd. Wire Co., Ltd. Products (Pinghu) Co., Ltd.
GPB MKCL KCH KWPF JYSF KSP
Michigan, Bangkok, Bangkok, Guangdong, Jiangyin, Pinghu,
Place
USA Thailand Thailand People's Republic of China People's Republic of China People's Republic of China
Secondary processing of Secondary processing and Secondary processing and Secondary processing and Secondary processing and Secondary processing
Nature of business CH & Bearing wire sales of cold finishing steel sales of CH wire sales of cold finishing steel bar sales of oil tempered wire and sales of CH wire
bar and CH wire
1995 1996
Establishment 1997 2004 2005 2007
(invested in March 2003) (invested in February 2002)
Capital US＄ 16.8 million THB 143 million THB 103 million JP￥ 725 million JP￥ 760 million JP￥ 1,200 million
Equity participation 20％ 27.5％ 30％ 60％ 60％ 50％
750 t/month
Nominal capacity 6,000 t/month 1,800 t/month 3,000 t/month for cold finishing bar 600 t/month 2,100 t/month
650 t/month for CH wire
Pickling & coating: 1 unit Combined drawing machine: Pickling & coating: 2 units Pickling & coating: 1 unit Pickling & coating: 1 unit Pickling & coating: 1 unit
Continuous furnace: 2 units Type Ⅰ 1 unit Shot-blasting: 1 unit Combined drawing machine Drawing machine with eddy STC furnaces: 2 units
STC furnace: 1 unit Combined drawing machine: Batch type furnace: 5 units with eddy current tester: 2 units current tester: 1 unit Drawing machine: 3 units
Main facility Drawing machine: 5 units Type Ⅱ 2 units STC furnaces: 4 units Drawing machine: 3 units oil tempered equipment: 1 unit
Eddy current tester: 1 unit Eddy current tester: 2 units Drawing machine: 9 units
Straightener: 1 unit
Cutting machine: 3 units

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 38

 companies and have implemented joint development 16,000 30.0
Other than Japanese auto
activities to produce special steel such as that used

Number of automobile production

14,000 Japanese auto
Ratio of Japanese auto 25.0

Ratio of Japanese auto (％)

in suspension springs. Kobe Steel also provides 12,000
technical guidance to establish the management 20.0

(×1,000 units)
10,000
know-how required for the supply bases providing
special steel to Japanese automakers in Europe. The 8,000 15.0

profile of ASCOMETAL is included in Table 1. 6,000

10.0
4,000
2.3 ASEAN 5.0
2,000

0 0.0
The annual production of automobiles in Thailand 2005 2006 2007 2008 2009
exceeded one million in 20051) (Fig. 4). The country, Year
Fig. 5 Automobile production in China
now dubbed the "Detroit of Asia", has become a
manufacturing base with a concentration of
automakers and parts manufacturers. Worthy of has not only remained, but continued to grow as a
special mention is that Japanese cars occupy more big consumer. Now it is fair to say that no industry
than 90% of the automobiles produced there and the can get along without China, which has overtaken
country is positioned as a huge base of operations Japan to become an economic super power that has
for exporting automobiles and auto parts to other the second largest GDP in the world.
countries in Southeast Asia, the Middle East and the China overtook the US as the world's largest auto-
Southern Hemisphere. making country in 20091) (Fig. 5). Major automakers
The eco-car project proposed by the Thai of the world have established manufacturing bases
government in 2007 was temporarily put on ice in there, with some manufacturers producing more in
the aftermath of the Lehman Shock; however, with China than they do in their own countries. New local
economic recovery, Japanese automakers, in particular, manufacturers are growing rapidly. Japanese and
began to announce plans to exploit the program and European component manufacturers are moving in
produce compact cars. Kobe Steel's contribution to one after another, increasing the demand for wire
the growth of the auto industry in Thailand includes rods and bars made of special steel.
the establishment, in the late 1990s, of Kobe CH Wire With this background, Kobe Steel has established
(Thailand) Co., Ltd., a production base for wire rods three bases for wire rod fabrication since 2004 and
for cold heading, and Mahajak Kyodo Co., Ltd., has built a structure for producing and supplying
which produces bright steel bars, a major material for wires for cold heading, bright steel bars for hot-coiled
hot-coiled suspension springs. The outlines of these suspension springs and oil-tempered wires for cold-
two companies are included in Table 2. coiled suspension springs. Kobe Wire Products
(Foshan) Co., Ltd. was established in Guangzhou, a
2.4 China major production center for Japanese automakers.
Jiangyin Sugita Fasten Spring Wire Co., Ltd. and
China is one of the few countries to have Kobe Special Steel Wire Products (Pinghu) Co., Ltd.
experienced ever-increasing economic growth, despite were established in the suburbs of Shanghai,
a brief stagnation after the Lehman Shock. The where there is a concentration of many component
country, which has been called "the world's factory," manufacturers. Kobe Steel thus has built a system for
the widespread supplying of secondary processing
2,000 products of wire rods to major auto production sites
Other than Japanese auto
1,800 in China. The outlines of these three companies are
Number of automobile production

Japanese auto
Ratio of Japanese auto
included in Table 2.
Ratio of Japanese auto (％)

1,600
100.0
1,400
(×1,000 units)

1,200 80.0 Conclusions

1,000
60.0
800 In response to the demand for customized steel
600 40.0 with consistent quality and secondary processing
400
20.0
products from Japanese automakers overseas, Kobe
200 Steel has established manufacturing bases in North
0 0.0 America, Europe, Thailand and China. These
2005 2006 2007 2008 2009
Year secondary processing bases for wire rods will
Fig. 4 Automobile production in Thailand continue to be used, despite the trend toward

39 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

reexamining standard steel with a view to cost References
reduction. Kobe Steel will expand its capacity to meet
demands in a timely manner and strive to build a 1) FOURIN, FOURIN Monthly Report on the Global Automotive
system to satisfy customers by working with its Industry (FY 2005 to 2009).
partners to improve the management skills of local
staffs in production and quality control.

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 40

 Development History of Wire Rods for Valve Springs
Nao YOSHIHARA＊1
＊1
Wire Rod & Bar Products Development Department, Research & Development Laboratory, Iron & Steel Business

To reduce fuel consumption, automotive valve springs Billet Hot-rolling Shaving Patenting Drawing Oil-tempering

need to be made lighter. Kobe Steel has developed high Eddy-current Testing Cold Coiling Stress Relieving Coil-end Grinding

strength steel for valve springs to meet this requirement. Nitriding

（     ） Shot Peening Strain Aging Setting
Furthermore, by continuously developing technology Fig. 2 Manufacturing process of valve spring
for the control of internal defects, such as non-metallic
inclusions, high strength valve spring steels with good
robust quality have been provided. This paper describes sag resistance under high stress. Until the late 1970s,
the development history of wire rods for valve springs valve springs had been made of hard-drawn wire
with high fatigue strength. of high carbon steel having a tensile strength of
about 1,700MPa. Then oil-tempered wires of
Introduction 1,900MPa grade began to be used. Nowadays, oil-
tempered wires of 2,100~2,200MPa grade are
Valve springs (Fig. 1) are used in the valve commonly used1), 2). Furthermore, nitriding treatment
actuation mechanisms of internal-combustion is being applied to the surfaces of wires to increase
engines. A valve spring serves to move an intake, or durability (Fig. 2).
an exhaust, valve according to the head-discharge This paper introduces Kobe Steel's history of
curve of a cam such that the valve is in contact with strengthening valve spring steel and newly developed
its seat to prevent compression leakage. In the materials.
meantime, a valve spring is required to impose
appropriate tension on the valve so as not to increase 1. Strengthening of valve spring steel at Kobe Steel
the friction loss of the valve operating system. The
environmental regulations for automobiles are 1.1 History of strengthening steel for valve springs
becoming more stringent in order to further reduce
CO2 emissions. To achieve this and to improve fuel Wires used for valve springs include piano wires
economy, valve springs are required to be lighter and (i.e., as-drawn high-carbon steel wires) and oil-
smaller. Such downsizing improves followability tempered wires (i.e., wires that are drawn, oil-
with cams, reduces the inertia weight of valve quenched and tempered).
operating systems, and helps to reduce the size of Until World War II, piano wires had been made
engines. However, as valve springs become lighter in Sweden. Kobe Steel started producing wire rods
and smaller, more stress is imposed on spring wires. made of high carbon steel in 1930. The company
In addition, the spring wires must withstand successfully developed wire rod for valve springs
repeated loading at a frequency of several thousand and inaugurated the production of piano wires for
cycles per minute for extended periods of time. valve springs in 1941. Around that time, the piano
Furthermore, aging degradation can cause shortening wires were first used for the valve springs of aircraft
of the spring, a phenomenon called "sag"; hence engines and then started to be used for the valve
valve springs are also required to have improved springs of automobile engines 3). In 1952, Kobe Steel
succeeded in developing a wire rod (KPR: Kobe
Piano Wire Rod) that was comparable with Swedish
steel 4).
After the war, oil-tempered wires were introduced
from the USA. In 1955, Kobe Steel began producing
oil-tempered wires of carbon steel and also of Cr-V.
Valve spring These wires gradually came to be used in automobile
engines 5).
Since around 1964, wires with excellent heat
resistance (SAE9254, JIS SWOSC-V), made of oil-
tempered steel containing Si and Cr, have been
Fig. 1 Valve spring in automobile engine used to increase fatigue strength and improve sag

41 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

resistance. This steel has become a de facto standard. strength by a factor of about 1.3.
Table 1 shows the chemical compositions of wire In the early 1990s, a new alloy, KHV10N,
rods used for valve springs, i.e., the compositions containing 2.0% of Si, was developed to increase
specified by SAE (SAE steels) and the ones developed resistance against temper softening. This steel has
by Kobe Steel. Fig. 3 depicts the development trend raised the tensile strength of oil-tempered wires to
of the wire rods. Kobe Steel's technologies include the 2,200MPa level. Combined with nitriding
techniques for reducing surface scratches and treatment and modified shot peening, KHV10N
decarburized layers, both of which can adversely has achieved a fatigue strength about 1.4 times
affect the fatigue strength of oil-tempered wires, higher than that of SAE9254.
and a method 6) for peeling the wire rod surface over To further improve the fatigue strength and sag
its entire length. In the early 1980s, the company resistance of KHV10N, larger amounts of Cr and V
developed and implemented technologies 7) for were added to make the crystal grains ultrafine.
evaluating and cleaning off harmful inclusions. The developed alloy, KHV12N, was commercialized
Adopting this technology to the SAE9254 steel has in 2006.
significantly improved the fatigue strength of valve Fig. 4 shows the ratio of high and super-high
springs. strength steels occupying the valve spring steels
Studies have been conducted aiming at optimizing produced by Kobe Steel. In recent years, the adoption
chemical compositions for higher strength. In the of high-strength steel has been increasing rapidly.
mid-1980s, a new composition, KHV7, was developed High-strength steel currently accounts for about 60
and implemented. This composition was based on percent of the valve spring steel shipped by Kobe
SAE9254 and has a higher content of C for increased Steel, and this rate is expected to increase.
tensile strength and an additional content of V,
which refines austenitic grain and improves softening 1.2 Means for strengthening
resistance. Oil-tempered wires of SAE9254 have a
tensile strength of about 1,900MPa, while those of Assuming no defect, the following relationship 8)
KHV7 have a tensile strength around 2,050MPa. The generally exists between the fatigue limit, σw, and
developed steel has an increased fatigue strength Vickers hardness, HV :
that is about 1.1 times higher than that of the σw＝1.6HV …………………………………………(1)
conventional SAE9254. Furthermore, applying wherein the units are in MPa for σw and kgf/mm2 for
nitriding treatment has increased the fatigue HV.

Table 1 Chemical compositions of wire rod for valve spring (mass％)

Steel grade C Si Mn Ni Cr V
KHV12N Ultra high tensile 0.60 2.15 0.45 0.20 1.75 0.27
KHV10N Super high tensile 0.58 2.00 0.85 0.30 0.95 0.10
KHV7 High tensile 0.62 1.45 0.60 − 0.60 0.12
SAE9254 SWOSC-V 0.55 1.40 0.65 − 0.65 −
SAE6150 SWOSV-V 0.50 0.25 0.80 − 0.95 0.20
SAE1070 SWO-V 0.70 0.25 0.80 − − −
Product ratio of high tensile valve spring steel (％)

160 100
(SAE9254＝100, 1×107 amplitude)

150 KHV10N
＋Nitriding 80
140 KHV12N
Fatigue strength

＋Nitriding
130 60
Regular
KHV7
120 ＋Nitriding
40
110
KHV7 High tensile
100 20
SAE9254
90
1975 1980 1985 1990 1995 2000 2005 2010 Super high tensile
0
Year 1985 1990 1995 2000 2005
Fig. 3 Trend of high strength steel grades for valve spring in Fig. 4 Product ratio of valve spring steel grades in KOBE
KOBE STEEL STEEL

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 42

 Valve springs for automobile engines are used 1.2.1 Nitriding treatment
over an extended period of time in a severe
environment of high temperature and high stress. Nitriding treatment is generally applied at a
In such an environment, a non-metallic inclusion temperature from 400 to 600℃. This treatment
larger than about 10μm can initiate fatigue fracture. hardens the surface layers of springs, increases
Murakami estimated the fatigue limit, σw, when compressive residual stress and significantly
internal defects such as inclusions were present, as improves fatigue strength. However, the treatment
follows 8). reduces internal hardness, making it difficult to
α
1.56(HV＋120) (1−R) satisfy both fatigue strength and sag resistance at the
σw ＝ ・ …………………(2),
(area)1/2 2 same time. To resolve this issue, oil-tempered wires
where R＝(σm−σa)/(σm＋σa), with softening resistance are very important.
α＝0.226＋HV×10−4,
area : Area of defect［μm2］, 1.2.2 Shot peening treatment
σm : Mean stress［MPa］,
σa : Stress amplitude［MPa］, Shot peening is effective in improving the
in which the units are in MPa for σw and kgf/mm2 fatigue strength of parts such as valve springs.
for HV This treatment increases the compressive residual
According to this formula, hardness must be stress and hardness of the treated surfaces so as to
increased and the defect size decreased to improve increase fatigue strength. Multi-stage shot peening
fatigue strength. A conventional method of improving has been employed for increasing the fatigue
fatigue strength is to increase the strength of wire by strength 10). Methods reported more recently include
oil tempering; however, when the tensile strength fine shot peening to improve residual stress 11), 12)
exceeds 1,800MPa, fracture is initiated at non- and enhanced shot peening for nano-sizing surface
metallic inclusions, causing the fatigue strength to crystals to improve fatigue strength 13). These shot
vary 1) as shown in Fig. 5. Thus, there is a limit to peening technologies should be combined with
improving fatigue strength by increasing the other surface modification technologies to achieve
tensile strength of wires. To overcome this issue, further strengthening.
several approaches have been undertaken, such as
increasing surface hardness by nitriding treatment 1.2.3 Inclusion control technology
and applying compressive residual stress by shot
peening. Compressive residual stress can be treated Non-metallic inclusions, which mainly consist of
as mean stress that reduces the effective stress. SiO2, can cause breakage of valve springs. These
Sag resistance, on the other hand, can effectively non-metallic inclusions are rendered harmless by
be improved by increasing the tensile strength lowering their melting point by controlling their
(internal hardness) of spring wires 9). compositions in the direction indicated by the
The following outlines the treatment and arrow 2) in Fig. 6.
technology for strengthening. To control the composition in the direction of

Initiation of Fracture Chemical Compositions (wt％)

Material Inclusion Surface Ｃ Si Mn Cr V
Si-Cr Steel ● ○ 0.55 1.40 0.70 0.70 − SiO2
Cr-V Steel ▲ △ 0.67 0.25 0.75 0.50 0.20
Plain Carbon Steel − □ 0.82 0.25 0.45 − −
10 90
950
00
17

20 80
1500

00
900 16
00
Fatigue strength (MPa)

30 15 70
1600
1700

0
140
850
40 60

800 Lower Melting Point 50

750 40
1400
1500
00

700 70 30
21

00
20

00

1800
19

00

650 80 20
18

00
17
210

150
140 0
2200

0
0
2300

160

0
1700

90 10
1900

600
240

2000
250

1,500 1,600 1,700 1,800 1,900 2,000 2,100 2,200

0
0

Tensile strength (MPa) 10 20 30
CaO (wt％) Al2O3
Fig. 5 Effect of tensile strength of steel wire for valve spring
9）
on fatigue strength Fig. 6 Composition of inclusion

43 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

the arrow, the processing parameters, such as slag Table 2 Properties of oil-tempered wire
basicity during secondary melting, must be Tensile strength Reduction of area γ grain
controlled more accurately. ［MPa］ ［％］ size No.

KHV12N 2,157 46.4 14.0

2. Characteristics of ultra-high strength wire rod
KHV10N 2,155 52.4 12.0
for valve springs

Kobe Steel developed, and commercialized in 2,300

2006, an ultra-high strength wire rod, KHV12N, 2,200
for valve springs; this wire rod has the world's

Tensile strength (MPa)

2,100
highest fatigue strength. This chapter introduces 2,000
the characteristics of KHV12N. 1,900
1,800
2.1 Concept of composition design 1,700
KHV12N
1,600 KHV10N
To improve the nitriding property and refine 1,500 SAE9254
crystal grains, larger amounts of Cr and V are added 1,400
than was the case with the previously developed 70

KHV10N. In addition, an increased amount of Si was 60

Reduction of area (％)

added to improve temper softening resistance in 50
order to prevent a reduction in internal hardness 40
during nitriding treatment and thus to enhance sag 30
resistance. 20 KHV12N
KHV10N
10 SAE9254
2.2 Characteristics of oil-tempered wires
0
300 350 400 450 500 550
Table 2 shows the mechanical properties and Annealing temperature (℃)

grain size numbers of oil-tempered wires. An ultrafine Fig. 7 Relationship between annealing temperature and
mechanical properties
grain structure (austenite grain size number 14) was
obtained for KHV12N. This oil-tempered wire was
Index of spring sag Index of spring fatigue life

subjected to low temperature annealing for 20 160

minutes for stress relieving after cooling. Fig. 7 140
shows the mechanical properties after annealing at 120
different temperatures. KHV12N exhibits improved
100
softening resistance and less strength reduction at
80
elevated temperatures even when compared with
120
another high-strength valve spring wire, KHV10N. 100
80
2.3 Fatigue characteristics of springs 60
40
20
Fig. 8 shows the fatigue and sag characteristics 0
SAE9254 KHV10N KHV12N
of springs made of three different wires. Nitrided
Fig. 8 Fatigue life and sag index of high-tensile valve spring
KHV12N exhibits a fatigue strength about 1.6 times
higher than that achieved by SAE9254. This enables
the spring weight to be almost halved. Fig. 9 shows
an example of a spring made of KHV12N.
Fig.10 shows the sag resistances of springs made
of various materials. KHV12N exhibits the highest
resistance against sag. The austenitic grain size
number turns out to correlate well with the tensile
strength of spring wires. The sag for KHV12N is
decreased by about 60% compared with that for
SAE9254, and by about 20% compared with that for
KHV10N. This is regarded as attributable to the SAE9254 KHV10N KHV12N
refinement of crystal grains. Fig.11 shows the Fig. 9 Comparison of valve spring height

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 44

 Index of residual strain (％) (SAE9254＝100％) 120
Index of residual strain (％)＝298−0.0546*[TS]−10*GS#
increased amount of Cr added in KHV12N improves
R＝0.995 the thermal stability of carbide, making the carbide
100  [TS]：tensile strength of spring wire (MPa)
 GS＃：austenitic grain size number SAE9254
more effective in preventing crystal grains from
coarsening. Furthermore, the increased addition of V
80
KHV7 increases the amount of carbide containing V,
60 KHV10N enabling the ultra-refinement represented by grain
size number 14. These are considered to be the
40 KHV12N reasons for the improved sag resistance.

20 Conclusions

0
0 20 40 60 80 100 120 Kobe Steel developed wire rods for valve springs
Index of residual strain calculated (％) earlier than any other company and has supplied
Fig.10 Sag resistance with a function of tensile strength and them to the market. The wire rods are characterized
austenitic grain size number by good robust quality and high strength, which
enable the steel to be used under high stress. After
8
2015, the production volume of hybrid vehicles (HV)
9 using both gasoline engines and electric motors is
expected to increase rapidly. Kobe Steel's high
Austenitic size number, GS#

10
strength wire rods for valve springs, including the
11 ultra-high strength series, are effective in downsizing
HV engines and in improving the fuel economy of
12
gasoline engines. Kobe Steel will strive to contribute
13 to the growth of the automobile industry and help to
resolve global environmental problems.
14

15 References
16
0.0 0.5 1.0 1.5 2.0 1) N. Ibaraki, R&D KOBE STEEL ENGINEERING REPORTS,
Cr content (％) Vol.50, No.3(2000), p.7.
Fig.11 Relationship between austenitic grain size number 2) N. Yoshihara et al., Transactions of Japan Society of Spring
and Cr content Engineers, No.45(2000), p.15.
3) Japan Spring Manufacturers Association, History of Springs
in Japan (1984), p.173.
4) K. Urakawa, R&D KOBE STEEL ENGINEERING REPORTS,
Vol.18, No.4(1968), p.29.
5) Japan Wire Production Association, SENZAI DOKUHON
(rev. 3), 1980, p.170.
6) Y. Nakamura et al., R&D KOBE STEEL ENGINEERING
REPORTS, Vol.24, No.1(1974), p.115.
7) T. Ohshiro et al., Stahl und Eisen, 109(1989), Nr.21, p.1011.
25μm 25μm
a) KHV10N : GS No.12 b) KHV12N : GS No.14
8) Y. Murakami, Effect of Micro-defects and inclusions，YOKENDO
Ltd. (1993), p.90.
Fig.12 Etched austenitic grain of ultra-high tensile valve
9) S. Suda et al., R&D KOBE STEEL ENGINEERING REPORTS,
spring
Vol.55, No.2(2005), p.22.
10) H. Toshino et al., Transactions of Japan Society of Spring
relationship between the amount of Cr added and Engineers, No.32(1987), p.31.
grain size. Fig.12 are micrographs of KHV12N and 11) Y. Yamada et al., SAE paper 2000-01-0791.
12) Y. Yamada et al., SAE paper 2003-01-1312.
KHV10N after grain boundary etching. The 13) H. Mano et al., Transactions of Japan Society for Spring
Research, Autumn 2003，p.23.

45 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

 Development Trends of Soft Magnetic Iron
Dr. Masamichi CHIBA＊1
＊1
Wire Rod & Bar Products Development Department, Research & Development Laboratory, Iron & Steel Business

Driven by advances in electronically controlled systems Plunger Coil winding

for automobiles, the demand is growing for soft
magnetic steels that can generate a large electromagnetic Core

force with low electric power. New steels with very low
carbon content have been developed in consideration Shaft

of their DC electromagnetic properties, their cold

forgeability, and machinability. This paper describes
the recent development trends of soft magnetic steel and
Solenoid cover
the advantages of our developed steel (ELCH2 series).
Fig. 1 Example of solenoid structure

Introduction
for hydraulic control. This is a typical example of the
Recently automobiles are being heavily equipped electromagnetic components used in automobiles.
with various systems operated electrically, electronically Soft magnetic materials are usually used for iron
and/or electromagnetically to ensure safety, comfort cores that form magnetic circuits and for the
and fuel economy. Such systems include computer housings, including covers.
controlled automatic transmissions, electronic power Early solenoid devices for automobiles were
steering (EPS) and anti-lock braking systems (ABS)1), 2). mainly used for controlling the oil flow by on-off
In the early years (1950s-1960s) when electric and switching. Thus, the availability and workability of
electronic devices started to be used in automobiles, iron core materials were more important than their
the main purpose was to improve and upgrade the magnetic characteristics. As a result, low carbon
performance of individual systems. Highly complicated steels such as SWRCH10A (JIS G 3507-1) and SUM23
vehicle control in recent years, however, is incomplete (JIS G 4804) were used for the iron cores.
without high-speed linkage and movements coordinated Recent solenoid devices, however, must be able to
among different systems. Hence electronically and/or control hydraulics swiftly and stably. To achieve this,
electromagnetically controlled components are becoming the on-off control is being replaced by linear control
increasingly important. in many cases.4)，5) For linear control, the control
Among these electronically and/or electromagnetically current must be proportional to the pulling force
controlled systems, components that utilize electromagnetic (electromagnetic force) of the iron cores, which
force incorporate iron cores in their coils to generate necessitates soft magnetic materials having low
magnetic fields. Many conventional cores consist of coercivity (low hysteresis) at high flux density. The
low carbon steel, containing about 0.1% of carbon, to latest iron core materials for linear solenoids are
improve the response to control signals and increase required to have magnetic characteristics better than
energy efficiency.3) With the remarkable advancement those of SUY-1 (JIS C 2504), making it difficult for
of electromagnetic control in recent years, it has low carbon steel, formerly used for on-off control, to
become essential for electromagnetic components to satisfy the required characteristics of magnetic flux
have even higher performance and lower power density and coercivity (Fig. 2).
consumption. In the mid-1980s, Kobe Steel worked to totally
This paper describes the trend of soft magnetic eliminate factors that can adversely affect the
materials, such as those for iron cores used in magnetic characteristics of soft magnetic materials
electromagnetic components, and introduces the and developed a soft magnetic material, ELCH2,
characteristics of the pure iron-based soft magnetic based on pure iron, which can achieve the best
materials in the ELCH2 (Extra Low Carbon Cold characteristics among JIS grade (JIS SUY-0) materials.6)
Heading Wire) series developed by Kobe Steel. The company has studied parts forming using a
forging process and parts function. Now this material
1. Transition of soft magnetic materials is being widely used as a soft magnetic material for
cold forging.
Fig. 1 depicts the structure of a solenoid device In the late 1990s, there was an increasing need for

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 46

 2.0 reducing Al and N content
Magnetic flux density (T, at 500A/m) JIS SUY-0 ELCH2 (Magnetic annealed)
Furthermore, the following aspects were
ELCH2S (Magnetic annealed)
1.5 SWRCH10A (Magnetic annealed) considered, taking into account the fact that metallic
ELCH2 (As rolled)
ELCH2S (As rolled) soft magnetic materials have the advantage of
SWRCH10A (As rolled) superior workability and high productivity:
1.0 SUM23L (As rolled)
iii) Improving cold forgeability by reducing Si
and adding Mn to render S harmless
0.5
iv) Improving the machinability of ELCH2S by
increasing S to disperse an appropriate
0.0
0 100 200 300 400 500 600 amount of MnS
Coercive force (A/m) Fig. 4 shows the relationship between the amount
Fig. 2 Comparison of magnetic properties of S added to ultra-low carbon steel and tool wear
width during turn cutting. It is expected that by
increasing the S content to about 0.025%, the tool
parts having more complex shapes and higher wear width will be decreased by half compared with
accuracy. This required machining after cold forging, the conventional pure iron based material containing
increasing the importance of the machinability of soft less than 0.010% of S.
magnetic materials. Pure iron based materials have However, as shown in Fig. 5, an excessive
excellent magnetic characteristics, but they are more addition of S deteriorates magnetic characteristics
difficult to machine than is carbon steel, due to significantly, while increasing the width of variation
factors such as chip handling during machining and in coercive force. Precipitation of FeS is observed at
decreased tool life. To resolve this issue, a small prior austenite grain boundaries.8) This indicates that
amount of Pb was added to the materials.7) However, the deterioration of magnetic characteristics is
Pb is a substance with a heavy environmental attributable to FeS, which decreases the compression
burden, hence forming parts without Pb has become ratio of ferrite phase determining magnetic moment
an important issue. and increases the resistance against domain wall
Kobe Steel studied techniques to improve migration. Because of this, the newly developed steel
machinability without sacrificing magnetic
characteristics and developed, in 2001, a new Pb-free 700
steel with improved machinability, ELCH2S.8) The Magnetic annealing：850℃×3h
600
commercial production of iron cores of large
Coercive force (A/m)

electromagnetic clutches started in 2004. The newly 500

developed material, as well as the previously 400

developed basic steel, is used for many automotive 300
electromagnetic systems with high-functionality.
200

2. Concept of developed steel 100

0
The magnetic characteristics of a soft magnetic 0.001 0.01 0.1 1.0
C content (mass%)
material depend on the magnetic moment of the
Fig. 3 C content dependence of coercive force
material, as well as its metallic structure, including
the size of crystal grains and precipitates. In the
60
case of a polycrystal, in particular, its magnetic
0.004％C-0.25％Mn (As rolled)
characteristics may deteriorate because of its crystal 50
Width of flank wear (μm)

grain boundaries and precipitates, both working as

pinning sites of domain walls.9) 40

Therefore, the magnetic characteristics of the

30
ELCH2 series have been improved along the Carbide AC200
following lines: Tool (Sumitomo Electric)
20
Cutting speed 260m/min
i) Making the structure a single phase of highly Feed rate 0.18mm/rev
10
pure ferrite to increase the magnetic moment Depth of cut 0.2mm
Cutting time 4min 20s
of the material, or reducing C content to no 0
higher than 0.01％ (Fig. 3) 0.0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04

ii) Decreasing the grain boundary area to reduce S content (mass％)

resistance against domain wall migration, or Fig. 4 Relation between S content and width of flank wear

47 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

 200 Table 1 Chemical composition of steels used in this study
0.004％C-0.25％Mn (mass％)
Magnetic annealing：850℃×3h
Steel C Si Mn P S
Coercive force (A/m)
150
ELCH2S 0.005 0.004 0.26 0.010 0.025

100 ELCH2 0.005 0.004 0.25 0.009 0.008

SWRCH10A 0.10 0.04 0.45 0.014 0.009

50 JIS SUY max. 0.03 max. 0.20 max. 0.50 max. 0.03 max. 0.03

0 3.3 Cold forgeability (deformability)

0.001 0.01 0.1
S content (mass％)

Fig. 5 S content dependence of coercive force

Deformability of the developed steel was evaluated
on specimens (φ20×30mm), each having a notch on
its lateral face. An upsetting test was conducted
with improved machinability, ELCH2S, is designed using a mechanical press while restraining the ends
to have an optimum Mn/S ratio, such that no FeS of each specimen. Deformability was determined
phase remains with increased S. from the critical upsetting ratio at which no cracking
occurs from the notch.
3. Specimens and experimental method
3.4 Machinability
Table 1 shows the chemical compositions of steels
that were studied: i.e., ELCH2 as a base composition, Machinability (i.e., burr height, chip disruption and
ELCH2S with improved machinability and wear width of cutting tool) was evaluated by a drill
SWRCH10A as a reference composition. These tested penetration test and cutting test using a carbide tool.
materials were melted in a converter furnace and
rolled into wire rods, each having a diameter of 3.5 Corrosion resistance
φ20mm. The rolled materials were evaluated for
their direct current magnetic characteristics, Corrosion resistance was evaluated by the salt
mechanical properties, machinability and corrosion spray test according to JIS Z 2371 (salt spray test
resistance. method), using 5％ brine water. After the salt spray,
the specimens were immersed in a 10％ solution of
3.1 Magnetic characteristics ammonium citrate at 70℃ for rust removal before
measuring the mass change.
Magnetic characteristics were evaluated on ring-
shaped specimens made of the respective test 4. Experiment results and discussion
materials. The evaluation was conducted according
to JIS C 2504 (Soft magnetic iron). An automatic 4.1 Magnetic characteristics
magnetometer (BHS-40, manufactured by Riken
Denshi Co., Ltd.) was used to obtain an initial Fig. 6 shows the relationship between the
magnetization curve and hysteresis curve, and these strength of the applied magnetic field and
curves were used to determine coercivity and magnetic magnetic flux density for ELCH2 series steels and
permeability. Also determined was the magnetic flux SWRCH10A. The tested materials had been annealed
density under each magnetic field. It should be noted in a vacuum at 850℃ for 3 hrs. This magnetic
that the sweep rate (200~250(A/ms)) of the applied annealing condition is widely used in the industry.
magnetic field during each measurement was set The ELCH2 series steels exhibit higher magnetic
within a range in which eddy-current loss does not flux density for a given applied magnetic field
affect the hysteresis curve. compared with the reference steel, showing
significant improvement for the magnetic field
3.2 Mechanical properties below 2,000A/m. The improvement effect seen in
flux density lessens with a higher magnetic field. This
Mechanical properties were evaluated by tensile is because, as magnetization approaches saturation, the
testing at room temperature. Tensile test pieces, 14 A magnetization mechanism shifts from domain wall
according to JIS Z 2201 (Tensile Test for Metallic migration to rotation magnetization, decreasing the
Materials), were prepared from the respective effect of domain wall migration governed by metallic
sample materials. structure.

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 48

 2.0 flux by about 65%.
In addition, as shown in Fig. 6, ELCH2 series
Magnetic field density (T)

1.5 JIS SUY-1 materials, as-rolled, satisfy magnetic characteristics

comparable with those achieved by SWRCH10A,
ELCH2 (Magnetic annealed)
1.0 ELCH2S (Magnetic annealed) magnetic annealed. Thus, the magnetic annealing
SWRCH10A (Magnetic annealed) step can be omitted for the parts currently made of
0.5
ELCH2 (As rolled) SWRCH10A, or of materials equivalent to it.
ELCH2S (As rolled)
SWRCH10A (As rolled)
4.2 Mechanical properties
0.0
0 1,000 2,000 3,000 4,000 5,000
Magnetic field strength (A/m) Table 3 shows the mechanical properties of the
Fig. 6 Magnetic field dependence of magnetic flux density developed steels, as-rolled. The as-rolled materials
have a tensile strength of about 300MPa and are
ELCH2 ELCH2S
comparable to spheroidized SWRCH10A. The tensile
FGc 4.0 FGc 4.0 strength is increased by working, such as wire
drawing, however, the softening that occurs during
magnetic annealing makes the ultimate tensile
strength about 230MPa regardless of the drawing
reduction.

4.3 Cold forgeability

0.1mm 0.1mm

Fig. 7 Microstructure of developed steel after magnetic Fig. 8 compares the critical upsetting ratios at
annealing which the notched specimens start to exhibit cracks.
The spheroidize annealed SWRCH10A exhibits a
Fig. 7 compares the microstructures after the critical upsetting ratio as high as 70%, while the
magnetic annealing of ELCH2 series steels. Each
material shows a clean structure of single phase Table 2 Magnetic properties of ELCH2 series (Magnetic
ferrite. It has been reported that, in the case of annealed)
ELCH2S with improved machinability, MnS may Magnetic field density (T) Coercive
Steel force
provide pinning points against the growth of crystal B100 B200 B300 B500 B1000 B4000 (A/m)
grain during magnetic annealing10); however, the ELCH2S 0.90 1.24 1.47 1.54 1.64 1.80 55.7
tested steels have almost the same size of ferrite
ELCH2 0.92 1.30 1.50 1.60 1.65 1.81 45.2
grains, and no adverse effect was observed. Thus, in
SUY-1 ≧0.60 ≧1.10 ≧1.20 ≧1.30 ≧1.45 ≧1.60 ≦80
the case of the ELCH2 series, both the base
SUY-0 ≧0.90 ≧1.15 ≧1.25 ≧1.35 ≧1.45 ≧1.60 ≦60
composition and modified steel with improved
machinability can have almost the same magnetic
characteristics, as long as the heat treatment Table 3 Mechanical properties of ELCH2 series
conditions are the same. Tensile Young Elongation Reduction
Table 2 shows typical magnetic characteristics of Steel strength modulus area
(MPa) (GPa) (％) (％)
the ELCH2 series. Both of the base compositions,
ELCH2 305 208 38.1 90.1
ELCH2 and ELCH2S with improved machinability,
ELCH2S 306 209 36.7 92.8
satisfy the superb magnetic characteristics achieved
by JIS SUY-0. Also, they can exert electromagnetic
force, comparable with that exerted by conventional 90
0.3mm
low carbon steel, with a smaller applied magnetic
Critical upset rate (％)

H h
field or smaller electric current, providing solutions 80
D
H/D＝1.5

for the improvement of component characteristics upset rate＝(H−h)/H×100

70
and power saving.
For example, assuming the magnetic flux density
60
required to actuate an electromagnetic device to be
1.6T, SWRCH10A requires a magnetic field of 50
1,200A/m, while the ELCH2 series materials require ELCH2 ELCH2S SWRCH10A SWRCH10A
(As rolled) (As rolled) (Spheroidizing (As rolled)
only about 400A/m, reducing the energy for annealed)
magnetomotive force required to produce magnetic Fig. 8 Critical upset rate

49 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

newly developed steels exhibit no cracking at an ELCH2 ELCH2S
upsetting ratio of 80%, even without softening
annealing, demonstrating excellent deformability.
Thus, ELCH2 series materials are effective in Chip
omitting process steps, such as heat treatment for
softening before cold forging, and reducing the 20mm 20mm
production cost of components.
Height
4.11mm 0.94mm
of burr
4.4 Machinability
Fig. 9 Comparison of chip and height of burr

(1) Drillability
Fig. 9 shows the chips and average burr height Table 4 Conditions for turning test
observed for the through holes drilled under the
conditions summarized in Table 4. In the case of Drill
Tool SKH straight drill
ELCH2S with improved machinability, the chips are
Sample
shorter and segmented with burr height significantly Cutting speed 30m/min
Height
reduced when compared with the base composition, Feed rate 0.20mm/rev of burr
ELCH2. The ELCH2S steel has MnS precipitated
Coolant Dry
and dispersed in its matrix. The precipitates behave
effectively as stress concentration sites in the chips,
improving the drillability. 0.15
(2) Turnability

Width of flank wear (mm)

Fig.10 shows the relationship between cutting
ELCH2
speed and flank wear width for ELCH2S with 0.10
ELCH2S
improved machinability. The steel was cut under the
conditions shown in Table 5. For a given cutting
0.05
length (800m), the wear width tends to decrease with
increased cutting speed. An increase in cutting speed
from 80m/min to 700m/min decreased the tool 0.00
0 100 200 300 400 500 600 700 800
wear to about 1/3 from 0.117mm to 0.038mm. The
Cutting speed (m/min)
improvement effect is significant in the cutting
Fig.10 Relation between cutting speed and width of flank
speed range up to about 250m/min. High-speed wear
cutting, with a cutting speed higher than 250m/min,
has turned out to be effective in reducing tool wear Table 5 Conditions for turning test
when cutting pure iron based material as the newly-
Tool Carbide AC200 (Sumitomo Electric)
developed steel. The following are considered to be
Cutting speed (m/min) 80, 150, 260, 350, 700
the reasons for reduced wear width during high-
Feed rate (mm/rev) 0.15
speed cutting11):
Depth of cut (mm) 0.2
i) The temperature increase associated with
Coolant Water-solube
cutting heat softens the work material,
decreasing the cutting resistance.
ii) A lower cutting speed suppresses the
generation of built-up edge on the cutting tool, SWRCH10A. Fig.11 shows the changes in corrosion
which promote tool wear. weight loss. The newly developed steel exhibits a
Fig.10 shows the tool wear width for the base corrosion weight loss about 40% smaller than that
composition, ELCH2, cut at a speed of 260m/min. of the reference steel, SWRCH10A, demonstrating
For the same cutting condition, the wear width for superior corrosion resistance compared with the
the developed steel is about half that of the reference material. The newly developed steel can
conventional steel, which verifies the effectiveness of be used without any problem for parts that have
the newly developed steel from the aspect of tool life. been made of conventional SWRCH10A or its
equivalent.
4.5 Corrosion resistance ELCH2 series materials are designed to have ultra-
low carbon content and a homogeneous structure
Salt spray tests were conducted on both the with suppressed precipitation to achieve improved
developed steel, ELCH2, and reference steel, magnetic characteristics. This is considered to

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 50

 100 of electromagnetic devices and to the reduction
Corrosion weight loss (g/m2 )
ELCH2 of their production costs
80 SWRCH10A
iii) A steel with improved machinability featured
60 by significantly reduced burrs and tool wear
during cutting, which reduces the work load
40 during cutting, compared with the base
composition, which highly prioritizes magnetic
20
characteristics
0
With the prevalence of hybrid electric vehicles
0 20 40 60 80 100 120 140 and electric vehicles, electromagnetically controlled
Time (h)
systems of new types are expected to be developed
Fig.11 Corrosion weight loss and applied. Kobe Steel strives to reduce the
environmental burden and thus to contribute to
suppress the generation of local cell in the corrosive society by providing pure iron based soft magnetic
environment, leading to excellent corrosion resistance. materials, the ELCH2 series, which make it possible
to improve the performance of electromagnetic
5. Application examples devices, save energy and reduce production costs.

ELCH2 series offer soft magnetic materials which References

are effective in improving and upgrading the
performance of electromagnetically controlled systems 1) A. HAMADA, "Mitsubishi Denki Giho" (in Japanese), 1987,
and in reducing the cost of parts production. Vol.61, No.8, pp.651-656.
2) T. KAWANOBE, "NAINENKIKAN" (in Japanese), 1987, Vol.26,
The materials have been employed for iron cores No.333, pp.230-241.
for hydraulic control solenoids, as well as for large 3) ASM Committee on Magnetically Soft Materials, Metals
iron cores for electromagnetic clutches, greatly Handbook Vol.1, American Society for metals, 1961, pp.785-
contributing to the reduction of power consumption 797.
and improvement of parts productivity12), 13). 4) N. FUKAYA, Journal of Society of Automotive Engineers of
Japan, 2004, Vol.58, No.4, p.48.
5) JTEKT Engineering Journal, 2007, No.1003.
Conclusions 6) M. CHIBA et al., R&D Kobe Steel Engineering Reports, 2002,
Vol.52, No.3, pp.66-69.
Pure iron based soft magnetic materials, the ELCH2 7) T. KATO et al., DENKI-SEIKO ELECTRIC FURNACE STEEL,
series, having excellent magnetic characteristics and 1967, Vol.38, No.1, pp.2-8.
8) M. CHIBA et al., R&D Kobe Steel Engineering Reports,
cold forgeability have been developed with the 2005, Vol.55, No.2, pp.18-21.
following features: 9) S. CHIKAKADO, Physics of Ferromagnetism, SHOKABO
i) Reduced power for generating electromagnetic PUBLISHING, 1963, 310p.
force, suppressing heat generation from 10) Y. ODA et al., NKK TECHNICAL REPORT, 2002, No.178,
components and reducing the load imposed pp.16-20
11) K. OZAKI et al., Proceedings of JSPE Semestrial Meeting, 2005,
on control circuits Vol.2005S, pp.593-594.
ii) Cold forgeability, enabling the production 12) R. KUNII et al., HONDA R&D Technical Review, 2004, Vol.16,
of complex shaped parts, which greatly No.2, pp.9-16.
contributes to the performance improvement 13) K. KUROKAWA et al., Journal of Society of Automotive
Engineers of Japan, 2005, Vol.59, No.8, pp.105-108.

51 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

 Influence of Ti Precipitate in Carburizing Steel Containing
Boron
Dr. Nariaki OKAMOTO＊1, Yosuke SHINDO＊1, Mutsuhisa NAGAHAMA＊2
＊1
Wire Rod & Bar Products Development Department, Research & Development Laboratory, Iron & Steel Business
＊2
Wire Rod & Bar Products Development Department, Research & Development Laboratory, Iron & Steel Business (currently with Wire Rod & Bar
Products Marketing & Technical Service)

When cold forged gear steel is carburized, fine 1. Considerations on steel for cold forged gears
precipitates of TiC prevent austenite grain from
coarsening. While preventing grain from coarsening, Steel for cold forged gears exploits the pinning
these precipitates increase the deformation resistance effect exerted by carbonitrides of elements such as
against cold forging due to precipitation hardening. Al, Nb and Ti to prevent the coarsening of crystal
This study evaluates the influence of TiC precipitates grains. To effectively exert this pinning effect, it is
on the deformation resistance, as well as their necessary to secure a large number density of
characteristics that work to prevent grain coarsening carbonitride precipitates during carburization by
during the carburization of a steel containing boron. suppressing Ostwald ripening, in which large
precipitates grow larger by absorbing adjacent
Introduction smaller precipitates. Since Ostwald ripening is a
phenomenon accompanying atomic diffusion, alloying
Cold forging has caught the attention of elements with low diffusion coefficients are preferable
automobile manufacturers in recent years because for effectively exerting the pinning effect 1). The
the process emits less CO2 and enables lower-cost diffusion coefficients of Al, Nb and Ti (Table 1)
production than conventional forming methods indicate that Ti is the most effective element in
such as cutting and hot forging. In particular, there exerting the pinning effect.
is an increasing need for the steel used to produce As shown in Fig. 1 2), the pinning effect can be
cold-forged gears. Cold forging of steel, however, is explained by a force acting to grow crystal grains
associated with a deformation resistance much (hereinafter referred to as "growth force") and a
higher than that associated with hot-forging. Because force exerted by precipitates and acting to suppress
of this, it is difficult to form complex shaped parts, grain growth (hereinafter referred to as "suppressive
such as gears, by cold forging. Thus steel for cold force"). The growth force is given by Formula (1)
forged gears is required to have a deformation according to the Gibbs-Thomson law:
resistance much lower than that of conventional gear
steel.
Table 1 Diffusion coefficient forming carbonitride
On the other hand, gears are carburized for
strengthening. If the gears are cold forged before Element
Diffusion coefficient
2 Precipitation
(m /sec)
carburizing, the strain energy induced by the cold
forging can cause crystal grain to coarsen during Al 3.0×10
−3
AlN
carburization, which may adversely affect the
−4
accuracy of the parts. To prevent this, steel for cold Nb 5.6×10 NbCN

forged gears contains alloying elements, such as Al, Ti 1.5×10

−5
TiC
Nb and Ti, which precipitate carbonitrides having a
"pinning effect". This pinning effect has long been
exploited to prevent crystal grain from coarsening. Deterrent force by a precipitation
The addition of these elements, however, increases ΔGpin

the deformation resistance. Thus, it is important to

exert the pinning effect effectively by carefully
controlling the precipitation state of the carbonitrides
while minimizing the amount of alloying elements
added. Grain growth force
With this background, a study was conducted on ΔG

the effect of the precipitation state of carbonitrides, as

reported in this paper. This study aims to develop a
steel that has a small deformation resistance, is Grain
effective in preventing crystal grain from coarsening
and is suitable for the cold forging of gears. Fig. 1 Schema of grain growth and pinning effect

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 52

 4σV coarsening, the carburizing steel containing Ti and B
ΔG ………………………………………(1)
   D was subjected to four kinds of heat treatment (i.e.,
wherein ΔG : growth force conditions ① to ④) to prepare steel samples having
   σ: grain boundary energy different precipitation states of TiC, as shown in
   V : molar volume Fig. 2 and Fig. 3.
   D : crystal grain diameter To study the deformation resistance, all the steel
On the other hand, the suppressive force is given samples (the carburizing steel containing Ti and B,
by Formula (2) according to the Zener-Smith model: heat-treated under conditions ① to ④, plus
πdσ SCM420H) were formed into solid cylinders, each
ΔGpin nV …………………………………(2)
   2 having a dimension of φ20×30mm. The cylindrical
wherein ΔGpin : suppressive force specimens were compressed with their ends
   d : diameter of precipitates confined with a compression rate of 75%. After this
   n : number density of precipitates per unit compression test, the specimens were carburized at
area 950℃ to investigate the characteristic of preventing
In the case where grain boundary growth is grain from coarsening. Observations to determine
suppressed, the growth force and suppressive grain size number were conducted in the area
force become equal (ΔG＝ΔGpin :), and thus the including an equivalent strain of 2.0, as shown in
relationship between the grain diameter and Fig. 4 (the area surrounded by a broken line in the
precipitates diameter is given by Formula (3). figure).
8
D …………………………………………(3)
   πnd
Formula (3) indicates that grain coarsening is Condition ① Condition ②
more effectively suppressed by precipitates grown in
a high number density and having large diameters.
On the other hand, the precipitates generated may
trigger the precipitation strengthening mechanism,
increasing the deformation resistance during cold TiC precipitation TiC precipitation
Density ：55.6particle/μm2 Density ：58.7particle/μm2
forging. The precipitation strengthening is given by Ave.size ：10.7nm Ave.size ：9.6nm
Formula (4) according to the Ashby-Orowan model 3) : Area ratio：0.50％ 100nm Area ratio：0.43％ 100nm
Condition ③ Condition ④
d
Δσ＝2.95 πn ln ……………………(4)
   2.5×10−4
where Δσ is the deformation resistance increment.
Formula (4) indicates that the deformation
resistance during cold forging increases with higher
TiC precipitation TiC precipitation
precipitate density and larger precipitate diameter. Density ：53.0particle/μm2 Density ：65.9particle/μm2
Ave.size ：10.6nm Ave.size ：8.6nm
In other words, properly controlling the precipitate Area ratio ：0.47％ 100nm Area ratio：0.38％ 100nm
density and grain size can suppress grain coarsening Fig. 2 Precipitation condition of TiC
during carburizing and prevents the deformation
resistance from increasing during cold forging.
40
30 Condition ① TiC(More than 20nm)
2. Experimental method 20
Ave. size ：20nm
Ave. density：2.4particle/μm2
10
Two types of steel, as shown in Table 2, were 0
TiC density (particle/μm2)

30 Condition ② TiC(More than 20nm)

used for the present experiment, i.e., carburizing Ave. size ：23nm
20 Ave. density：5.6particle/μm2
steel containing Ti and B, and SCM420H, which is 10
generally used for gears. In order to clarify the effect 0
Condition ③ TiC(More than 20nm)
of precipitation states on the deformation resistance 30
Ave. size ：20nm
and on the characteristic of preventing grain from 20 Ave. density：1.6particle/μm2
10
0
Condition ④ TiC(More than 20nm)
Table 2 Chemical composition of specimens 30
Ave. size ：27nm
20 Ave. density：2.4particle/μm2
Chemical composition (mass％)
Steel 10
C Si Mn Cr Mo Ti B 0
3 5 8 10 12 15 18 20 22 25 30
Carburizingsteel containg B 0.17 0.06 0.44 1.38 − 0.04 add TiC size (nm)
SCM420H 0.20 0.18 0.83 1.12 0.17 − − Fig. 3 Condition of TiC distribution

53 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

(mm)
2.8 size numbers of specimens carburized after a
10
9 2.4 compression test. These micrographs indicate that
8
2.0
the grain size numbers of the top-most surface layers
7
1.6
all fall around No.12 regardless of which of the ①
6
5 Observation
1.2
to ④ conditions applies; they show no difference. It
area
4
0.8
should be noted however that, at a depth deeper
3
2 0.4
than 400μm from the surface layer, conditions ①, ③
1
0.0
and ④ (unlike condition ②) yielded coarser grains
0 with grain size numbers no greater than No. 3,
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
(mm) as observed in the area surrounded by circles in the
Fig. 4 Position of investigation for size of grains
figures. This is despite the fact that conditions ① to
④ all yield an average grain size number of around
3. Experimental results No.8.

Fig. 5 shows the deformation resistance of materials 4. Discussions

softened under conditions ① to ④. Almost no
difference was found in the deformation resistance 4.1 Effect of precipitation state on deformation
for carburizing steels containing Ti and B, treated resistance
under conditions ① to ④. (Their hardness is about
16% lower than that of SCM420H.) The results Formula (4) indicates that the deformation
indicate that the difference in the precipitation state resistance during cold forging is affected by the grain
of TiC, investigated in this experiment, does not size and density of precipitates. The materials,
significantly affect the deformation resistance. treated under conditions ① to ④, have precipitates
Fig. 6 depicts micrographs showing the grain (TiC) with different diameters and densities, as
shown in Fig. 2 and Fig. 3; however, they resulted
700 in almost no significant difference, with their
deformation resistances falling around 542 to
Deformation resistance (MPa)

649 544MPa, as shown in Fig. 5. The effect of the

650
precipitation state on deformation resistance is
16％DOWN
600 calculated by applying the values of the density and
average size, shown in Fig. 2, to Formula (4). As
550 543 544 542 544 shown in Table 3, the increments in deformation
resistance for the precipitation state of conditions ①
500 to ④ are calculated to be 143 to 150MPa, showing
hardly any difference, just as in the case of the
450 measured values. Therefore, the precipitates from
① ② ③ ④ SCM420H
conditions ① to ④ are regarded as having almost no
Fig. 5 Deformation resistance of each specimen
effect on the deformation resistance.
Condition ① m) Condition ②
(μm) m)
(μm)
0 0 4.2 TiC effect of preventing crystal grain coarsening
100 100
200 200
Large grain As shown in Fig. 6, all the conditions, ① to ④,
300 300
400 400 resulted in surfaces with a grain size number of
500 500 No.12 (grain size: about 6μm), with hardly any
Surface GhNo. ：12.0 600 Surface GhNo. ：12.0 600
Average GhNo.：8.0 Average GhNo.：8.0 differences. These fine crystal grains are attributable
Large GhNo. ：3.0
m)
(μm)
Large GhNo. ：−
m)
(μm)
to the pinning effect of TiC. According to Formula
Condition ③ Condition ④
0 0 (3), grain size can be evaluated by precipitate
100 100
Large grain
200 200
diameter and density. Assuming that TiC, observed
300 300 before carburization, remains unchanged after
Large grain
400 400 carburizing, Formula (3) yields grain sizes of 4.3 to
500 500
600 600
4.5μm for conditions ① to ④, using the values
Surface GhNo. ：12.0
Average GhNo.：8.5
Surface GhNo. ：12.0
Average GhNo.：8.0
shown in Fig. 2. These values agree with the
Large GhNo. ：3.0 Large GhNo. ：2.0 observed values, showing hardly any differences.
Fig. 6 Grain size of carburized specimen after compressing These results indicate that the majority of TiC
test precipitates effectively exert a pinning effect.

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 54

 On the other hand, for conditions ①, ③ and ④, dissolved into the matrix to form a solid solution,
coarse crystal grains with a grain number of 8 Formula (3), provided with their average diameter
(grain size: about 22μm), were observed at a depth and number density, yields grain sizes for conditions
deeper than 400μm (Fig. 6). These coarse grains are ① to ④, as shown in Fig. 8. Fig. 8 indicates that
attributable to the less effective pinning effect condition ②, which caused no crystal grain
exerted by smaller TiC precipitates, which are coarsening yielded a measured grain size that agrees
energetically less stable than larger ones, and well with the calculated value. On the other hand,
dissolve into the solid matrix during carburization, conditions ①, ③ and ④, which caused crystal grain
decreasing the number density of the precipitates. coarsening, resulted in calculated values greater than
The solid solubility of these precipitates can 40μm (grain number, No.6). These results consistently
generally be explained by solubility curves such as elucidate the fact that, at a depth deeper than 400μm
the one shown in Fig. 7 4). In Fig. 7, dissolved Ti from the surface, TiC precipitates smaller than 20nm
becomes more stable as the concentrations of Ti and dissolve into the matrix to form a solid solution
C approach the lower left corner, and TiC becomes during carburization, losing their pinning effect and
more stable as the concentrations approach the upper leaving only TiC greater than 20nm exerting the
right corner, with respect to the border curve for pinning effect.
the carburizing temperature of 950℃. Thus it is From the above, the behaviors of TiC and crystal
considered that the carburized steel containing B grains, during carburization in the B-containing
has a lower C concentration in its interior than in carburizing steel, are classified into two cases as
its surface layer, making TiC less stable. It is also shown in Fig. 9, i.e., Case I, where a small number
considered that, in the interior of the carburizing of TiC precipitates greater than 20nm exist; and Case
steel containing B, smaller TiC precipitates, which II, where a large number of TiC precipitates greater
are energetically less stable, dissolve into the solid than 20nm exist. In Case I, the surface layer has a C
matrix, leaving relatively larger TiC precipitates in concentration high enough to maintain TiC
the matrix during carburization. Thus, it is possible
to quantitatively evaluate the crystal grain 200
GhNo. 2
Calculation
coarsening from the condition of the relatively 180
Measurement (maximum grain size)
160
larger TiC precipitates observed before carburization.
Grain size (μm)

140 GhNo. 3 GhNo. 3

Assuming that the TiC precipitates larger than 120
20nm, as shown in Fig. 3, remain effective in exerting 100
80
a pinning effect during carburization without being 60 Standard of Grain
40 grain growth growth
Table 3 Deformation resistance and calculation of the 20 No grain
influence by precipitation hardening 0 growth
Condition Condition Condition Condition
Measurement (MPa) Calculation (MPa) ① ② ③ ④
Heat history Increase of deformation by Fig. 8 Comparison of measurement value and calculation in
Deformation resistance
precipitation grain size around 400μm from surface
Condition ① 543 146
Condition ② 544 146 Case Ⅰ Minor amount of TiC more than 20nm (Condition ①, ③, ④)
Condition ③ 542 143 High density
Before Carburizing After Carburizing Hard solute
Condition ④ 544 150

C density
Solubility curve：log[Ti][C]＝−4230/(T＋273)＋3.23 TiC solute
0.06
TiC
Stable Low density
Solubility curve at 950℃ Grain Growth Easy solute
1,300℃ TiC under 20nm
TiC more than 20nm Solute TiC
0.04
Ti (mass％)

Composition Composition Case Ⅱ Major amount of TiC more than 20nm (Condition ②)
at inside at surface
Before Carburizing After Carburizing High density
Hard solute
Stoichiometric Composition 1,200℃
0.02 [Ti]＝3.99[C]
C density
1,100℃ TiC solute
Sol. Ti
950℃ 1,000℃
Stable
900℃ Low density
0.00
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 TiC more than 20nm Solute TiC Easy solute
C (mass％) TiC under 20nm No grain growth

Fig. 7 Solubility Curve of TiC Fig. 9 Schema of relation of TiC and grain size by carburization

55 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

precipitates stably regardless of their sizes, which (1) Conditions ① to ④, applied for varying
prevents the crystal grain from coarsening. precipitation states without changing the
Meanwhile, in the interior, the C concentration composition, resulted in no significant difference
becomes low, causing TiC precipitates smaller than in the deformation resistance.
20nm to dissolve into the matrix to form a solid (2) Conditions ① to ④ yielded different grain size
solution, which decreases the number of particles distributions after carburization following
contributing to the pinning and causes crystal grain compression. Crystal grain coarsening was
coarsening. On the other hand, in Case II, the surface prevented effectively in the surface layer under
layer has a C concentration high enough to maintain all of the conditions; however, only condition ②
TiC precipitates stably regardless of their sizes, was effective in preventing the grain from
which prevents crystal grain coarsening; this is the coarsening at a depth of around 400μm from the
same as in Case I. However, in the interior, C surface layer.
concentration becomes so low that TiC precipitates (3) For conditions ①, ③ and ④, grain coarsening
smaller than 20nm dissolve into the matrix to form a occurred at a depth of about 400μm from the
solid solution. Meanwhile, TiC precipitates larger surface. This is attributable to the lower density of
than 20nm remain undissolved, ensuring the number TiC precipitates larger than 20nm, compared
of particles contributing to the pinning, which serves with the case of condition ②.
to prevent crystal grain coarsening.
References
Conclusions
1) E. Fujita, Kinzoku-Butsuri, AGUNE GIJUTSU CENTER
A study was conducted with the aim of (2004), p.659.
2) T. Nishizawa et al., Thermodynamics of microstructure，Japan
developing a steel for cold forged gears that has Institute of Metals (2005), p.140.
decreased deformation resistance and the effect of 3) W. C. Leslie, Ferrous material, Maruzen (1985), p.213.
preventing crystal grain coarsening. The following 4) H. Ohtani et al., Proceedings of Japan-Canada Seminar on
results were obtained: Secondary Steelmaking (1985), p.1.

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 56

 Low Alloy Steel for Fracture Splitting Connecting Rod
Akihiro MATSUGASAKO＊1
＊1
Wire Rod & Bar Products Development Department, Research & Development Laboratory, Iron & Steel Business

To reduce vanadium content in the steel for fracture Conventional process

Hot forging Dowel pin
splitting connecting rods, the effect of titanium addition Crude Dowel
on the splitting property was investigated. The results processing processing
indicate that Ti addition effectively reduces the impact Finish
processing
value of the steel. This is considered to be attributable Hot forging
Crude Dowel
to the decrease in the volume fraction of ferrite and processing processing

in the precipitation strengthening of the ferrite. Notch

Fracturing process
Furthermore, Ti was also found to have no adverse
Hot forging
effect on machinability. Based on these results, a new Notch&
Crude Finish
Fracturing
processing processing
steel was developed for fracture splitting connecting processing
rods.
Fig. 1 Manufacturing process of connecting rods
Introduction
Steel for fracture splitting connecting rod must
Among the characteristics required for be split with minimum deformation (fracture
automobiles, "low environmental burden" and splittability), in addition to possessing the
"safety" are two key features gaining increasing characteristics required for conventional steel used
recognition. The environment, in particular, is in connecting rods. In Europe, a steel, C70S6,
receiving much public attention, and automakers are according to the DIN standards, is widely used.
actively moving forward to develop low-emission This steel has a microstructure that is almost entirely
and fuel-efficient cars 1), 2). occupied by pearlite to ensure fracture splittability.
Automobiles must be manufactured at a low Kobe Steel also developed a steel for fracture
cost while achieving the above objectives. To splitting connecting rod with high-strength. The
reduce the cost, low alloy steel is being pursued, as steel ensures fracture splittability by the increased
is a simplified manufacturing process. With this addition of P and V, as well as spheroidized MnS 7).
background, a method of production according to To avoid the risk of relying on scarce resources,
the fracture splitting process was developed to Kobe Steel pursued a technique for improving the
reduce the cost of manufacturing connecting rods, fracture splittability with less addition of V and has
which are automobile engine parts 3). The conventional newly developed a steel for fracture splitting
process for making a connecting rod includes forging connecting rod based on newly acquired knowledge.
and rough-machining the rod portion and cap portion This paper outlines that development.
separately, providing these portions with dowels
and coupling them before finishing. On the other 1. Study on alternative means for low alloying
hand, according to the fracture splitting process, the
rod and cap portions are forged and rough-machined Although C70S6 steel contains no alloying element
as a unified body and notches are machined on the with a high availability risk, it suffers from a lack
body along a plane, which later becomes the of machinability because it contains a large amount
coupling surface. In the subsequent step, the forged of hard cementite. On the other hand, non-heat-
body is fractured along these notches to split the treatment steel, consisting of ferrite and pearlite, has
rod and cap portions, and finish machining is fewer problems with machinability; however, the
performed on these split portions (Fig. 1). This toughness of its proeutectoid ferrite (hereinafter
production method improves the material yield of simply referred to as "ferrite") must be reduced to
the forged body and eliminates the need for improve its fracture splittability. This is clear from
processing cap portions, dowel pins and the holes. the fact that reduced impact strength improves
Since the fracture splitting method was implemented fracture splittability 7). A method is being developed
in the 1990s, it has prevailed in Europe and US, and to reduce its impact resistance by adding V and/or
an increasing number of Japanese manufacturers P to increase the strength and thus to embrittle the
have introduced this method 3) - 6). ferrite.

57 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

 Because Vanadium is produced by a limited Notch
φ8.3 r：0.2
number of countries and is a scarce resource with depth：0.5
availability risk 8), an alternative element has been

Direction of extension
sought. Thus, Kobe Steel has focused on Titanium,
which is considered to have a large output with less
43
risk of unavailability, and which forms carbide as φ

65
in the case of V. The company conducted a study
on the effect of Ti addition on impact resistance
and machinability to determine the feasibility of
substituting Ti for V.

22
2. Experimental procedure 65
(unit：mm)
2.1 Materials tested Fig. 2 Fracture splitting test piece

Table 1 shows the chemical compositions of the

materials tested. Four types of steels were tested, i.e.,
two types of steels (Steel A and Steel B), with or
without Ti, to study the effect of Ti, a steel (Steel C)
L1 L2
equivalent to C70S6 as a reference for evaluating
fracture splittability, and a steel (Steel D), equivalent Fracture
splitting
to the conventional steel used for connecting rod,
as a reference for evaluating machinability. The
steels were melted in a laboratory scale vacuum Before fracture splitting After fracture splitting
melting furnace, and hot forged at 1,200℃ into a Hole diameter increase＝L2−L1
predetermined shape before the testing. Fig. 3 Measuring method of hole diameter increase

2.2 Impact test

2.4 Fracture splitting test
To study the effect of Ti on impact strength, Steel
A and Steel B were subjected to a Charpy impact The fracture splittability was evaluated by a
test. A hot forged cylinder having a diameter of fracture splitting test. A hot forged round bar with a
φ30mm was hot pressed in a diametrical direction diameter of φ65mm was hot pressed in a diametrical
to produce a plate having a thickness of 15mm, from direction to produce a plate having a thickness of
which a U-notched specimen was prepared. 25mm. The plate was machined into a fracture
splitting specimen as shown in Fig. 2. The notches
2.3 Machinability test were provided in a direction vertical to the forging
direction such that the fracture split would occur at
Drill life was evaluated to study the machinability. positions that simulate actual connecting rods. The
A hot forged cylinder having a diameter of φ80mm fracture splittability was evaluated by the change in
was provided as a work material whose cross-section the hole diameters before and after fracture splitting.
was subjected to drilling for drill life evaluation. The In other words, as shown in Fig. 3, the hole diameter
tool used was a high-speed steel drill with a diameter was measured before the fracture splitting and was
of φ10mm. The test was conducted under dry measured again after the split parts were fastened by
conditions with a feed rate of 0.21mm per revolution. bolts. The difference in the hole diameter was taken
The total cutting length until the drill broke was as an index of fracture splittability.
taken as the drill life.
3. Experimental results and discussions
Table 1 Chemical compositions of samples（mass％）
3.1 Impact test
Sample C Si Mn P S Cr V Ti
Steel A 0.38 0.25 1.07 0.051 0.050 0.18 0.170 0.051
Fig. 4 shows the results of the Charpy impact
Steel B 0.38 0.24 1.08 0.049 0.050 0.18 0.160 −
tests. It has turned out that the addition of Ti
Steel C 0.70 0.23 0.54 0.011 0.058 0.11 − − decreases the impact resistance significantly. The
Steel D 0.41 0.26 1.06 0.019 0.060 0.23 0.099 − fracture surfaces and microstructures of impact

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 58

 notch
60 dimple
Charpy impact value
50

Fractured surface
40
(J/cm2)

30
20
100μm 100μm
10
a) Steel A b) Steel B
0 Fig. 5 SEM images of fractured surface
Steel A Steel B
Fig. 4 Result of Charpy impact tests
Ferrite fraction：20％ Ferrite fraction：29％
Hardness：294.7HV Hardness：272.0HV
specimens Steel A and Steel B were observed in detail
to elucidate the phenomena that occurred during the
impact test.

3.1.1 Fracture surface observation

100μm 100μm

a) Steel A b) Steel B
Fig. 5 shows the SEM micrographs of fracture
Fig. 6 Microstructures and hardness of Steel A and B
surfaces. Steel A exhibits cleavage throughout almost
the entire surface without any of the dimpling that is
A：Austenite B：Bainite
a feature of ductile fracture. The partially observed F：Ferrite M：Martensite
grain boundaries indicate that the fracture, to put it P：Pearlite
1,000
more exactly, is caused by quasi-cleavage. Steel B, on
the other hand, exhibits dimples throughout almost Ac3 820deg
800
the entire surface of the specimen at the notch bottom A
Ac1 740deg

where the initial crack was introduced. It should be F

TEMP (deg)

600 P
noted that the cleavage fracture surface is located
more than about 300μm inside the fracture surface. 400
B

This indicates that the decrease in impact

resistance brought about by the addition of Ti is 200
M

attributable to the suppression of ductile fracture Hardness

367 302
588 522 362 305 324 309 287
accompanying dimples. These dimples are generated 0
(HV)
2 3
by the plastic deformation of matrix surrounding 1 10 10 10 104 105
TIME (s)
inclusions, and in a ferrite-pearlite structure, the a) Steel A
plastic deformation is mainly governed by ferrite 9). 1,000
Thus it is highly possible that the addition of Ti
serves to embrittle the ferrite. 800
Ac3 820deg

Ac1 740deg
A
F
3.1.2 Microstructure
TEMP (deg)

600
P

Microstructures were observed to study the B

400
change in the ferrite. Fig. 6 shows the results of the
M
study. The addition of Ti has turned out to 200
decrease the ferrite fraction from 29% to 20%, Hardness
316 328
273 167
(HV) 551 551 302 344 295
while increasing the hardness from 272.0HV to 0
2 3
1 10 10 10 104 105
294.7HV. The decreased ferrite fraction raises the TIME (s)
ductile-brittle transition temperature 9), lowers the b) Steel B

upper shelf energy 10) and lowers the impact Fig. 7 Continuous cooling transformation diagram
resistance value. These effects are considered to be
the cause of the lowered impact value. behavior during hot forging. The results are shown
Now, a detailed study was carried out to in Fig. 7. The results indicate that the addition of
elucidate the cause of the ferrite fraction's being Ti delays the initiation of diffusion transformation
decreased by the addition of Ti. Continuous cooling (ferrite transformation), shifting the curves toward
transformation (CCT) diagrams were prepared to longer times and lower temperatures.
study the transformation behaviors. The heating The following discusses the reason why the Ti
temperature was set to 1,200℃ to simulate the addition delays the initiation of ferrite transformation.

59 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

Firstly, Ti is an element having a strong affinity adverse effect of Ti addition on machinability. Steel
with C, forming carbide. Another carbide forming D is the steel equivalent of conventional connecting
element, V, has been known to decrease the diffusion rod.
rate of C 11). Thus, Ti is assumed to decrease the
diffusion rate of C, as in the case of V. The decreased 3.3 Fracture splitting test
diffusion rate is considered to suppress diffusion
transformations such as ferrite transformation and Based on the knowledge obtained in Sections 3.1
pearlite transformation. and 3.2, compositions were determined for low alloy
Next, a detailed study was conducted on the steel for fracture splitting connecting rod. The low
increase of hardness. The hardness increase is alloy steel has a strength (proof strength of about
attributable both to the precipitation hardening effect 600MPa) equivalent to that of the conventional
of TiC and the decrease in the ferrite fraction. To connecting rod steel (Steel D). Table 2 shows a
verify the effect of the ferrite fraction on the sample composition of the newly developed steel.
hardness change, a mixing rule was adapted for This steel contains Ti in about the same amount as
hardness. The mixing rule for hardness is given by in Steel A, in which the effect of Ti addition has
the following relationship 12): been sufficiently confirmed. It also contains V in
HV ＝ fα1/3HVα＋ （1− fα1/3） HVp …………………(1) about the same amount as in Steel D to achieve the
in which HV represents overall hardness, fα same level of strength. The newly developed steel
represents the ferrite fraction, HVα represents ferrite contains V in an amount reduced by about 60%
hardness and HVp represents pearlite hardness. compared with the conventional steel made by Kobe
The effect of the ferrite fraction was calculated for Steel for fracture splitting connecting rod.
the ferrite fractions shown in Fig. 6, assuming HVα＝ Fracture splitting tests were conducted on the five
235 and HVp＝295. According to the calculation, a 9 steels; i.e., Steels A-D and the newly developed
percent decrease in ferrite fraction increases the steel. Fig. 9 shows the test results. The newly
hardness by about 5HV. Thus, in the total hardness
increment of about 23HV, shown in Fig. 6, 1,000
A
approximately 18HV is attributable to precipitation B
hardening by TiC. This precipitation hardening is C
considered to have caused the ferrite to harden and 100 D
Drill life (cm)

embrittle.

3.1.3 Embrittlement mechanisms by Ti addition 10

From the above, the following is proposed to be

the mechanism of Ti addition decreasing the impact 1
0 10 20 30 40 50 60 70 80
resistance. Cutting speed (m/min)
The addition of Ti serves i) to decrease the ferrite Fig. 8 Results of drill life test
fraction, raise the ductile-brittle transition temperature
and lower the upper shelf energy; and ii) to cause Table 2 Chemical compositions of developed steel
precipitation hardening, to increase the strength of （mass%）
ferrite and to decrease the plastic deformability. Sample C Si Mn P S Cr V Ti
These effects cause cleavage fracture to occur Developed steel 0.37 0.25 1.12 0.047 0.062 0.29 0.106 0.043
while suppressing the ductile fracture that
accompanies dimples, leading to reduced impact
resistance. Particularly in this study, the ferrite Steel A

fraction had a significant effect, which can be

Steel B
exploited in improving the fracture splittability by
reducing the impact resistance value. Steel C

3.2 Machinability test Steel D

Developed
Fig. 8 shows the results of the machinability test. steel
The machinability of Steel C (C70S6 equivalent) was 0.0 0.1 0.2 0.3 0.4
dramatically degraded, while Steel A, Steel B and Hole diameter increase (mm)

Steel D exhibit almost equal tool lives, indicating no Fig. 9 Hole diameter increases

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 60

 developed steel and Steel A, both containing Ti, References
exhibit a fracture splittability that significantly
exceeds that of Steel B and Steel D without Ti 1) Edition by Society of Automotive Engineers of Japan, Inc.,
addition and is better than that of Steel C, the C70S6 Material Technology for Automobiles, 5 (1996), p.2, Asakura
Publishing Co., Ltd.
equivalent. The results confirm that the addition of 2) Edition by Society of Automotive Engineers of Japan, Inc.,
Ti enables the improvement of fracture splittability Automobile Technology Handbook, 1 (2008), p.1, Society of
with a decreased addition of V. Automotive Engineers of Japan.
3) Z. Gu et al., The International Journal of Advanced Manufacturing
Conclusions Technology, Vol.25, No.9-10 (2005), pp.883-887.
4) T. Kubota et al., YAMAHA MOTOR TECHNICAL REVIEW,
No.39 (2005), pp.80-87.
With the aim of developing a low alloy steel for 5) S. Koganemaru et al., Proc. of Annual Conference of JSAE, No.
fracture splitting connecting rod, a new technique 20-06 (2006), p.7.
was devised to improve the fracture splittability, in 6) N. Kyotoku, Society of Automotive Engineers of Japan,
which Ti is substituted for V. It has turned out that Symposium, Newly Developed Engine, No.19-07 (2008), p.19.
7) G. Anan, et al., R＆D KOBE STEEL ENGINEERING REPORTS,
the addition of Ti decreases the ferrite fraction, Vol.56, No.3 (2006), pp.44-47.
induces the precipitation hardening of the ferrite, 8) Japan Oil, Gas and Metals National Corporation, Mineral
facilitates cleavage fracture and reduces the impact Resources Material Flow 2009, (2009), pp.147-152.
resistance value. Furthermore, the Ti addition has 9) T. Araki, Materials Engineering for Steel (1970), p.102,
been found to ensure favorable fracture splittability Maruzen.
10) S. Okamoto : Steel Materials (1960), p.167, Corona Publishing
without sacrificing machinability. Based on the new Co., Ltd.
knowledge, a new steel, containing much less V, was 11) S. N. Tewari et al., Journal of Materials Science, Vol.17, No.6
developed for fracture splitting connecting rods. (1982), pp.1639-1648.
12) T. Gladman et al., Journal of The Iron and Steel Institute,
Vol.210 (1972), pp.916-930.

61 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

 Production Technology of Wire Rod for High Tensile
Strength Steel Cord
Kazuhiko KIRIHARA＊1
＊1
Wire Rod Production Department, Kakogawa Works, Iron & Steel Business

The wire rod for steel cord is designed for high tensile
strength. The steel cord is typically used to reduce tire Silicon ingot
Feed Saw wire
weight. For several years, such steel cord has also been
used as saw wire to cut silicon ingots. The saw wire is
more highly tensile and has a smaller diameter.
Therefore, an appropriate wire rod is required for
manufacturing such saw wire. This report describes the Return
technology to control non-metallic inclusions and the
wire rod with superior drawingability needed for Fig. 1 Schematic of cutting of silicon ingot
manufacturing the high strength wire used for steel cord
and saw wire. means for suppressing the strain on the cut surface,
it is effective to increase the strength of the saw wires.
Introduction Here again, high strength wires are demanded, as in
the case of steel cord.
Steel cord, with its excellent strength and This paper describes some approaches taken to
resilience, is used in various industrial applications reduce the diameter of steel cord and to simplify the
such as reinforcing materials for tires and conveyor production process. Also described is the projected
belts. future of the wire rods used for steel cord.
The steel cord used for automobile tire
reinforcement is a strand of fine wires of ultrafine 1. Qualities required of steel cord
drawn steel, called filaments. The downside of steel
cord is its specific strength, lower than that of other Table 1 shows typical chemical compositions of
reinforcing materials consisting of chemical fibers wire rods produced by Kobe Steel for steel cords.
such as rayon, nylon and polyester; this is a Fig. 2 shows the trend toward increasing the strength
downside that increases the weight of tires. On the of steel cord 1). The tensile strength of a filament with
other hand, steel cord has the advantages of a high a diameter of 0.20mm was 2,800MPa in the 1970s,
modulus of rigidity, as well as excellent thermal 3,300MPa in the 1980s, and reached a high-strength
conductivity, which prolongs the life of tires of 3,600MPa in the early 1990s and 4,000MPa in the
significantly and provides automobiles with superior late 1990s 2), 3). Steel cord strengthened by wire drawing
traveling performance and driving stability.
Table 1 Chemical compositions of wire rod for steel cord
With the proliferation of automobiles in newly (mass％)
developing countries, the use of steel cord is Steel grade C Si Mn P S Cr
expected to increase dramatically. To meet the KSC72 0.70-0.75 0.15-0.30 0.40-0.60 ≦0.020 ≦0.020 ≦0.05
market needs for lighter tires, improved fuel KSC82 0.80-0.85 0.15-0.30 0.40-0.60 ≦0.020 ≦0.020 ≦0.05
KSC90 0.88-0.93 0.15-0.30 0.40-0.60 ≦0.020 ≦0.020 ≦0.05
economy and a simplified manufacturing process,
KSC92-E 0.90-0.95 0.10-0.25 0.30-0.50 ≦0.020 ≦0.020 0.10-0.30
there is a demand for steel cord that can be produced KSC105-E 1.02-1.07 0.15-0.30 0.20-0.40 ≦0.020 ≦0.020 0.10-0.30
in larger quantities and with even higher strength.
5,000
Meanwhile, the market for solar energy generation
Wire dia.：φ0.2mm
has been rapidly growing in the last several years,
Tensile strength (MPa)

4,500
since it emits no CO2 during power generation and
is in accord with the global effort to reduce the 4,000
environmental burden. The silicon wafers used for 4,000MPa
3,300MPa
0.92%[C]＋Cr
solar panels are sliced from silicon ingots using saw 3,500 0.82％[C]
2,800MPa 3,600MPa
wires (Fig. 1). To improve the yield of the material to 0.72％[C] 0.82％[C]
3,000 0.92％[C]
be sliced, a saw wire must have a fine diameter. A
sliced wafer, on the other hand, is required to have 2,500
1970 1980 1990 2000
a cut surface with strain suppressed in order to Year
achieve high efficiency in power generation. As a Fig. 2 Trend of high tensile strength of tire cord

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 62

 Wire rod Coarse Patenting Intermediate Final patenting
(φ5.5mm) drawing (φ3.0mm) drawing (φ0.8∼φ1.6mm)

Brass Wet Filament Stranding Tire cord

plating drawing (φ0.15∼φ0.38mm)

Fig. 3 Manufacturing process of tire cord2）

follows the microstructure of pearlitic steel which 4,400

Tensile strength (MPa)

has a layered structure of ferrite and cementite. Steel 4,200

cord has the highest strength among the mass 4,000

produced materials in practical use. To make it 3,800

stronger, the carbon content in the wire rod has 3,600

been increased from 0.7％ C (hypoeutectoid steel) 3,400

to 0.8％C (eutectoid steel) and then to 0.9％C 3,200

(hypereutectoid steel). 3,000

0.18 0.16 0.14 0.12 0.10 0.08
Fig. 3 depicts a typical manufacturing process Wire diameter (mm)
of steel cord used for tire reinforcement 2). Fig. 4 Tensile strength of saw wire
A hot-rolled wire rod with a diameter of φ5.5mm 3
is processed according to the steps of dry wire
Wire Breakage Index

Wire dia.：φ0.25mm
drawing, intermediate patenting, brass plating and
2
wet wire drawing, up to which point both the steel
cord and saw wire are produced by almost the same
process. The filaments that constitute steel cord have 1
extremely small diameters of φ0.15 -φ0.38mm and
are subjected to strong torsional stress during the 0
wire-stranding step after the wet wire drawing. 2,600 2,800 3,000 3,200 3,400 3,600
Tensile Strength (MPa)
This can cause the filament to break, decrease
Fig. 5 Relationship between tensile strength of filament and
productivity and degrade the filament quality. wire breakage index during stranding process
Therefore, stringent quality is required throughout
the entire length of each filament.
A saw wire, on the other hand, is used as a solid
filament and does not include a stranding step in its
manufacturing process. The wire has an ultrafine
diameter of φ0.08 -φ0.20mm to improve the yield
of the material to be cut. In addition, the wire is
repeatedly subjected to bending stress and tensile
stress during cutting. Thus the tensile strength of a
saw wire can reach as high as 4,000MPa (Fig. 4).
Fig. 5 shows the relationship between the tensile
strength of a filament for steel cord and the wire Fig. 6 Fracture surface of steel cord
breakage frequency (index) during the stranding
step, showing a rapid increase in the breakage rods for steel cord, and non-metallic inclusions must
frequency as the strength increases.3) The cause of be further reduced to prevent the wire breakage
the breakage, attributable to the wire rod, includes associated with increased strength.
surface scratches, centerline segregation and
inclusions. In particular, an inclusion as small as 2. Challenge to attain zero-breakage
several tens of microns can not only become the
starting point for wire breakage (Fig. 6), but also An inclusion that can cause wire breakage mainly
affect the fatigue characteristics 4). Thus, the cleanness consists of hard alumina, in particular, corundum
of the steel definitely affects the quality of wire (Al2O3) and spinel (MgO・Al2O3), which are unlikely

63 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

to elongate and likely to fracture during hot rolling vicinity of the eutectic line between anorthite and
and cold working 5) - 7). pseudowollastonite and in the area surrounding
Alumina can either crystallize out from molten the primary phase of spessatite. Kobe Steel has
steel, or originate from refractories. To avoid this, succeeded in preventing hard alumina from being
various measures have been implemented, including crystalized out and in rendering the inclusions
the prevention of Al from mixing into the molten harmless by establishing a production technology for
steel, modification of the slag refining method and accurately controlling slag so that it meets a target
improvement of refractories. composition during slag refining.
(1) Prevention of Al adulteration (3) Improvement of refractories
The concentration of dissolved Al in molten steel Refractories are indispensable for treating molten
in equilibrium with a target composition is as low steel at high temperatures. However, refractories can
as several ppm. Therefore, the adulteration of Al, cause wire breakage. Kobe Steel has established a
e.g., from alloy iron, must be controlled or technology for using refractories as materials for
restricted. Kobe Steel not only uses alloys with receiving molten steel so that the molten steel is
restricted Al concentration, but also has established maintained at maximum cleanness while retaining
a technology for suppressing Al in molten steel to a their strength and corrosion resistance.
very low concentration by using a slag composition
controlling technology during slag refining, as
described later.
(2) Slag refining
Kimura et al. studied the fracture behavior of
alumina, zirconia, zircon and silica, all of which are
hard and have high melting points, during hot
rolling and cold wire drawing 9). The micrographs in
Fig. 7 show the fracture behaviors of alumina and
silica in hot rolled wire rods (φ5.5mm) and in the
subsequent cold drawn wires 9). No significant
difference appears in the fracture behavior of the
hot rolled materials; however, in the cold drawn
wire, silica, having lower compressive strength, is
fractured into smaller pieces. Inclusions that are
crystallized out from molten steel, or are adulterated
from refractories, should be modified into glassy
compositions with a lower melting point such that
they can be fractured more easily.
Inclusions in wire rods for steel cord are
roughly classified into a CaO-SiO2-Al2O3 type having
its origin in slag and a MnO-SiO2-Al2O3 type having
its origin in deoxidation products. Fig. 8 shows the
phase diagrams of the CaO-SiO2-Al2O3 and MnO-
SiO2-Al2O3 systems. In both systems, the composition
range in which the melting point becomes low
enough for the compound to elongate during hot
rolling, and thus become harmless, lies in the Fig. 8 Aim of chemical compositions of inclusions

Steel size φ5.5mm φ4.8mm φ1.2mm

Alumina

Silica

Direction of rolling and drawing 20μm

Fig. 7 Typical examples of change in shape of oxide inclusions appeared on longitudinal section of steel wire during drawing 9)

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 64

 3. High drawability - Development of new product exhibits no delamination at a higher wire drawing
strain, demonstrating its excellent drawability.
As described above, there is an increasing
demand for stronger steel cord. Research and 4. Future prospects
development in pursuit of the ultimate high
strength are being done on steel cord with strength With its excellent characteristics, steel cord for
exceeding 4,000MPa. tires has been playing an important role in traveling
On the other hand, for the purpose of reducing performance with its features including driving
the production costs incurred by cord manufacturers, stability and weight reduction, which improves
new methods are being developed for direct drawing fuel economy. As the world focuses on the reduction
and for improving the life of dies. Direct drawing of the global environmental burden, exhaust gas
does away with the intermediate patenting that regulation in automobiles will become even more
exists in the primary wire drawing of conventional stringent, requiring higher strength steel cord that
materials. enables a further weight reduction in tires.
Kobe Steel has been pursuing wire rods with Increasing the strength of steel cord is an effective
higher strength as well as with excellent drawability, means for reducing the weight of tires; however,
which enables direct drawing and improves the life increasing the strength of a material increases its
of dies. defect sensitivity. Thus, further reduction is required
An index for evaluating the drawability of a wire for defects such as non-metallic inclusions, segregation
rod is the reduction of the area of the wire rod, in and surface scratches.
which a higher reduction of area is more suitable for On the other hand, competition in the manufacturing
wire drawing. It is also indispensable to decrease the of steel cord has become increasingly intense, requiring
tensile strength of wire rod by increasing the interval materials that contribute to cost reduction and
distance of lamellar pearlite, which enlarges the productivity improvement.
critical zone for wire drawing and increases the In addition, the market for solar energy
life of dies. generation will continue to grow, which will further
In general, the tensile strength and reduction of enhance competition among saw wire manufacturers
area of a wire rod are in a proportional relationship, and ingot cutting manufacturers. Cuttability, assured
in which the lower the strength is, the lower the by high strength, and cutting yield, assured by small
reduction of area becomes. By combining controlled wire diameter, require a wire rod with high strength
rolling and controlled cooling, Kobe Steel has and fewer defects that can cause wire breakage, as in
developed a wire rod having excellent drawability the case of steel cord application.
with reduced tensile strength, while maintaining a
high reduction of area. Conclusions
A torsion test is performed to evaluate the
soundness of wire, in which vertical cracking, called Kobe Steel will strive to develop materials
delamination, occurs when a wire embrittled by the meeting the market needs and thus contribute to the
wire drawing is twisted. growth of industries related to tires, automobiles and
Fig. 9 compares the relationship between the solar energy generation.
true strain and the critical strain for delamination
generation in a conventional steel to that of the References
newly developed steel. The newly developed steel
1) Y. OKI, Bulletin of The Iron and Steel Institute of Japan, Vol.8,
2,000 No.9(2003), pp.627-632.
Developed wire rod 2) T. Minamida et al., R&D Kobe Steel Engineering Reports,
Pick up tensile strength (MPa)

Conventional wire rod Vol.50, No.3(2000), p.32.

1,500 3) Y. Yamada et al., R&D Kobe Steel Engineering Reports, Vol.36,
No.4(1986), p.71.
4) M. Tomioka et al., R&D Kobe Steel Engineering Reports,
1,000
Vol.23, No.3(1975), p.39.
Delamination 5) A. Yoshimochi et al., Wire Journal Int., Sep(1983), p.224.
500 (Developed wire rod) 6) H. SATO et al., SEITETSU KENKYU, Vol.320 Jan., 1986, p.35.
7) E. Stampa et al., Wire Journal Int. Mar(1987), p.44.
Delamination
(Conventional wire rod)
8) T. Mimura, 183th Nishiyama Kinen Gijutu Koza, 2004, pp.11-12.
0 9) S. Kimura et al., Tetsu-to-Hagane, Vol.88, No.11(2002), pp.755-
0.0 1.0 2.0 3.0 4.0 762.
Strain (ε)
Fig. 9 Relationship between strain and delamination

65 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

 Characteristics of Brittle Crack Arrest Steel Plate for
Large Heat-input Welding for Large Container Ships
Masahito KANEKO＊1, Dr. Tokutaka TANI＊1
＊1
Plate Products Development Department, Research & Development Laboratory, Iron & Steel Business

Once it happens, brittle fracture in the hatch coaming initiated by chance, it is essential that a back-up
parts around the deck openings of container ships brittle crack arresting function be included in the
causes serious structural damage that could potentially steel plates 2). Many studies have been conducted on
result in both fatalities and environmental damage. the crack-arrestability of steels. It is reported that a
With this in mind, ships are designed and constructed so brittle fracture test, performed on a model test body
as to ensure that brittle crack does not occur. Further, simulating a T joint for a hatch side coaming and
if by chance it does occur, having a back-up function upper deck, gave a result indicating that a steel plate
for arresting brittle crack included in the steel plate is having a thickness of 60mm can serve as an
essential. This report describes the characteristics of effective crack arrester if it has a Kca value (brittle
KE36 class brittle crack arrest plates. Improvement in crack propagation-arrest toughness) no smaller than
brittle crack arrestability was achieved by the 6,000N/mm1.5 at a test temperature of -10℃ 3, 4).
refinement of crystal grains, which is a result of strictly However, only a few reports refer to the methods
temperature-controlled TMCP (Thermo Mechanical for producing such a heavy thickness in steel plates
Control Process). for hull structures, i.e., a plate thicker than 50mm
and having high arrestability with a Kca value
Introduction (-10℃) exceeding 6,000N/mm1.5.
In this development work, heavy-thickness steel
With the recent increase in the volume of marine plates were control rolled under optimum conditions
transport, container ships are becoming larger. Now, with stringent roll control in the temperature zones
very large ships that can carry more than ten that respectively cause recrystallization and non-
thousand containers have been built. A container ship recrystallization. As a result, a technique was
has a structure with a large opening on its upper deck established for producing a steel plate with a high
and its hull girder constructed with an open cross- arrestability with a Kca value (-10℃) exceeding
section. This requires container ships to have the 6,000N/mm1.5. This paper outlines an overview of
highest longitudinal strength among large merchant the production technique and introduces the
ships. In order to ensure longitudinal strength while characteristics of the newly developed steel plate.
upsizing its hull, each container ship has a hatch side
coaming surrounding its deck and an upper deck, 1. Development target
both thickly built adopting steel plates no thinner
than 50mm1). Table 1 shows the target properties to be
The interior of a thick plate, however, is in a plane achieved by this development work. The target
strain state with the plastic region decreased in size. mechanical properties for the base metal and
As a result, a stress greater than its yield stress is welding joint are to meet the requirements of the
generated, and cracks propagate more easily. Brittle Nippon Kaiji Kyokai (NK) standard, KE36.
fractures, once occurring in the hatch coaming part The arrest characteristics of the base metal aim
around the deck openings of a container ship, can to satisfy the minimum brittle crack-arrest toughness
cause serious structural damage, with the potential at the test temperature of -10℃, Kca (-10℃), to be
for both fatalities and environmental damage. With no lower than 6,000N/mm1.5, according to the
this background, ships are designed and constructed "Guidelines on Brittle Crack Arrest Design" proposed
so as to ensure that brittle crack does not initiate. by Class NK 4).
Furthermore, in case a brittle crack should be

Table 1 Target properties

Base metal properties Arrestabilty Properties of welded joints
Thickness
Grade YP TS EL vE−40 Kca(−10℃) Welding TS vE−20
(mm) 1.5
(MPa) (MPa) (％) (J) (N/mm ) method (MPa) (J)
≧34(Ave.) 1pass ≧34(Ave.)
KE36 60 ≧355 490∼620 ≧21 ≧6,000 ＊ 490∼620
≧24(Each) EGW ≧24(Each)
＊
EGW (Electrogas welding)

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 66

 2. Development concept technique involves rolling continuously across the
high temperature regions from the recrystallization
Improvement of arrest characteristics is known temperature zone to non-recrystallization temperature
to be achieved by several techniques including: 1) zone. This conventional method is modified by
refining the grain size of the surface layer to 1 - 3μm 5); incorporating steel plate cooling during the rolling to
and 2) balancing the grain refining and the deformation allow tighter temperature control. A study was
texture of ferrite 6). conducted on the rolling in the low temperature
This development work aims to establish a region at a non-recrystallization temperature
technique for producing a steel plate having immediately above the Ar3 transformation point.
arrestability with a Kca value (-10 ℃) exceeding The temperature is considered to facilitate grain
6,000N/mm1.5 by fully exploiting the capacity of the refinement by effectively introducing strain
existing facilities. (nucleation sites) to austenite grains.
As shown in Fig. 1, Kca is reported to correlate
with the toughness of a base metal at t/4 portion 3. Features of newly developed steel
(fracture surface transition temperature vTrs)7). If
there is no slip deformation, cracks in a steel material 3.1 Chemical composition and mechanical properties
generally propagate along a crystal plane with a low of base metal
surface energy before causing fracture8). Therefore,
the toughness of the base metal can be effectively The chemical composition of the newly developed
improved by increasing the grain boundaries having steel is shown in Table 2. To ensure toughness at low
misorientation angles greater than 15 degrees. Such temperatures in the heat affected zone (HAZ) caused
grain boundaries serve as resistance against brittle by high heat input welding, the C content is limited
cracks 9). Grain boundaries with misorientation angles to 0.08% in order to prevent the toughness
greater than 15 degrees are here referred to as "high deterioration caused by island-shaped martensite.
angle grain boundaries." It is important to refine For this purpose, Ceq is maintained as low as 0.34%.
grains surrounded by high angle boundaries to A small amount of Ti is added to prevent the
improve the toughness of the base metals, because toughness deterioration caused by the coarsening of
such refinement is considered to result in the prior austenite grain. Furthermore, a small amount
improvement of the arrest characteristics (Kca≧ of Nb, which expands the non-recrystallization
6,000N/mm1.5). In the case of heavy-thickness steel temperature zone, is added in order to promote the
plates, however, the applicable rolling reduction is so refinement of the grains surrounded by high angle
limited and the temperature deviation in the boundaries.
thickness direction is so large that the conventional The relation between the size of grains
rolling technique will not allow rolling in the surrounded by high angle boundaries and Kca (-10℃)
required temperature zones, making refinement was studied under various TMCP conditions. The
difficult. result is shown in Fig. 2. As predicted, Kca is
To resolve this issue, a technique called thermo
mechanical control process (TMCP) was introduced. Table 2 Chemical compositions of developed steel 
(mass％)
This technique enables an adequate control of the C Si Mn Ti Others Ceq
rolling reduction in the recrystallization and non- Developed
0.08 0.12 1.55 0.012 Nb, B, Ca 0.34
Steel
recrystallization temperature zones 10). The technique
Ceq＝C＋Mn/6＋
（Cr＋Mo＋V）/5＋（Cu＋Ni）
/15
was fine tuned for the rolling temperature regions and
rolling reduction. More specifically, the conventional 100,000

9,000
Kca at −10℃ (N/mm1.5 )

8,000
Kca at −10℃ (N/mm1.5 )

7,000
6,000
10,000
5,000
4,000 6,000N/mm1.5
3,000
2,000
1,000 1,000
0 0 5 10 15 20
−100 −80 −60 −40 −20 0 Grain size (μm)
vTrs (℃)
Fig. 2 Relationship between high angle grain size and Kca at
Fig. 1 Relationship between Kca at −10℃ and vTrs (t/ 4) −10℃

67 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

improved by refining grains surrounded by high steel tested. The newly developed steel plate exhibits
angle boundaries. By making these grains no larger the mechanical properties that satisfy the target with
than 10μm, the target of Kca (-10℃)≧6,000N/mm1.5 vE−40, greater than 340J, sufficiently exceeding the
was achieved. target value (34J). As shown in Fig. 4, the developed
Now, a study was conducted to establish a steel plate exhibits a favorable fracture surface
production technique for decreasing the grain size transition temperature of lower than -90℃.
to under 10μm while maintaining high angle
boundaries to ensure arrest characteristics. Fig. 3 3.2 Brittle crack propagation-arrest characteristics
shows micrographs and electron backscattering
diffraction (EBSD) patterns mapping the crystal Brittle crack propagation-arrest characteristics
orientations at grain boundaries of newly developed were evaluated by Kca values, calculated from the
and conventional steels 11). The grains mapped are results obtained by an ESSO test 5) specified by the
those surrounded by high angle boundaries with Guidelines on Brittle Crack Arrest Design. Fig. 5
crystal misorientation greater than 15 degrees. The shows the result of the ESSO test. By extrapolating
conventional steel has a structure mainly consisting the fitted line, the developed steel is expected to have
of upper bainite. By introducing an adequate amount Kca exceeding 8,000N/mm1.5 at the lowest use
of strain in a low temperature range between the temperature (-10 ℃), which sufficiently satisfies the
recrystallization temperature and non-recrystallization minimum brittle crack arrest toughness value Kca
temperature, the newly developed steel is changed so (-10 ℃)≧6,000N/mm1.5 specified by the Guidelines
as to have a structure mainly consisting of polygonal on Brittle Crack Arrest Design.
ferrite in which grains surrounded by high angle Fig. 6 shows an example of the fracture surface,
boundaries with crystal misorientation greater than 100
15 degrees are refined. This is attributable to the 90 Longitudinal
strain introduced at a low temperature between the 80 Transverse
recrystallization temperature and non-recrystallization 70
Crystallinity (％)

temperature, which serves to form ferrite nuclei in 60

the austenite grains and to promote the generation of 50
polygonal ferrite12). 40
Table 3 shows the base metal properties of the 30
20
Microstructure EBSD 10
0
−140 −120 −100 −80 −60 −40 −20 0

400
Developed
350
Absorbed energy (J)

300

250

200

150
Conventional 100

50
50μm
0
50μm −140 −120 −100 −80 −60 −40 −20 0
※Position：t/4 Test temperature (℃)
Fig. 3 Microstructure and grain boundary map with EBSD Fig. 4 Charpy transition curve of developed steel plate

Table 3 Mechanical properties of developed steels

Base metal properties
Thickness ＊1 ＊1 ＊1
(mm) YP TS EL vE−40 vTrs
＊2
(MPa) (MPa) (%) (J) (℃)
Developed 340 (Ave.)
60 425 538 31 −100
Steel 338, 332, 349 (Each)
Conventional 231 (Ave.)
60 499 615 23 −60
Steel 235, 230, 229 (Each)
KE36 Target ≧34 (Ave.)
60 ≧355 490∼620 ≧21 −
properties ≧24 (Each)
＊1 ＊2
Round tensile specimen：NK14A  Charpy test specimen：NKU4

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 68

 10,000 Table 4 Conditions of EGW
Developed steel
Conventional steel Root Number Welding Welding Welding Heat
Thickness Groove Welding
｡ gap of current voltage speed input
(mm) angle( ) consumable
6,000N/mm1.5 (mm) passes (A) (V) (cm/min) (kJ/cm)
Kca (N/mm1.5 )

Wire：DWS-1LG
60 20 8 (φ1.6mm) 1 400 43 2.3 449
Shielding gas：CO2
3,000

Table 5 Mechanical properties of EGW welded joint

−10℃ −20℃ −30℃ −40℃ Properties of welded joints
＊2
Thickness vE−20min (J)
1,000 TS
＊1
Broken
3.6 3.8 4.0 4.2 4.4 (mm)
(MPa) location Position Depo Bond Bond＋
1/T×10 3 (K−1) 1mm
Fig. 5 Results of ESSO test Surface 113 208 199
60 583 HAZ t/2 91 204 203
Arrested
Back 108 207 182
KE36 490 ≧34(Ave.)
−
Target properties ∼620 ≧24(Each)
＊1 ＊2
Round tensile specimen：NKU2A   Charpy test specimen：NKU4

Conclusions

An adequate amount of strain was introduced

Fig. 6 Fracture surface of specimen after ESSO test at a low temperature in the range between the
recrystallization temperature and non-recrystallization
in which brittle crack is arrested as a result of temperature to produce a microstructure mainly
toughness improvement associated with the consisting of fine polygonal ferrite. As a result,
temperature gradient. It is considered that, as a grains surrounded by high angle boundaries were
shear-lip extends from the surface layer to the successfully refined in a heavy- thickness steel plate
interior (t/8 - t/4) of the steel plate, the driving having a thickness exceeding 50mm.
force of brittle fracture becomes smaller than the The results satisfy the minimum brittle crack
kinetic fracture toughness value at the center of plate arrest toughness value Kca (-10 ℃)≧6,000N/mm1.5
thickness. This serves to arrest the brittle crack 8). specified in the Guidelines on Brittle Crack Arrest
The developed steel exhibits fracture surface Design. The developed steel responds to the need for
having a shear-lip extending to t/8 - t/4. The plastic larger and safer container vessels, which continue to
deformation at the shear-lip is considered to absorb be built. The demand for the newly developed steel is
energy for crack propagation, improving the arrest expected to grow rapidly.
characteristics.
References
3.3 Characteristics of high-heat-input welding joint
1) H. SHIRAKIHARA, Technology and Challenge in Application of
High-heat-input welding is performed for Steel Material in Shipbuilding, NISHIYAMA KINEN GIJUTSU
KOZA, 2007-06.
assembling the hatch side coaming and upper deck 2) Y. YAMAGUCHI, KANRIN (in Japanese), No.3, 2005, p.70.
of a container ship to assure high construction 3) E. TAMURA, CAMP-ISIJ, Vol.20, 2007, p.469.
efficiency. To simulate the work done at these 4) Nippon Kaiji Kyokai, Guidelines on Brittle Crack Arrest
portions, electro-gas welding (EGW) with a single Design, 2009.
electrode was performed with high heat input. 5) T. Ishikawa, Tetsu-to-Hagane, Vol.85, No.7(1999), pp.544-551.
6) K. NISHIMURA, JFE GIHO, No.18, 2007, p.19.
The welding conditions for the developed steel 7) E. TAMURA, CAMP-ISIJ, Vol.22, 2009, p.1315.
plate are shown in Table 4 and the welding joint 8) S. AIHARA, NISHIYAMA KINEN GIJUTSU KOZA, 177th,
characteristics are shown in Table 5. The welding 2002, pp.159-160.
was performed with a high heat input of 450kJ/cm. 9) M. KANEKO, ISOPE, 2010.
The joint strength satisfies the target value. In 10) M. KANEKO et al., CAMP-ISIJ, Vol.22, 2009, p.1315.
11) S. SUZUKI, Materia Japan, Vol.40, No.7(2001), p.612.
addition, positive results were obtained regarding 12) The Iron and Steel Institute of Japan, Recent study on the
the joint toughness: a Charpy absorbed energy of 34J banitic structure and its transformation behavior of (ultra) low
or more was achieved at all notch positions in the carbon steel (in Japanese), 1994-07-30.
V-notch Charpy impact test at a test temperature
of -20℃.

69 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

 Effect of Steel Toughness on Brittle Crack Arrest
Behavior of T-weld Joint Structure Using Thick Plates
Dr. Eiichi TAMURA＊1, Dr. Tomokazu NAKAGAWA＊2, Kazuyuki TSUTSUMI＊2, Naohiro FURUKAWA＊3
＊1
Materials Research Laboratory, Technical Development Group
＊2
Mechanical Engineering Research Laboratory, Technical Development Group
＊3
Plate Products Marketing & Technical Service, Iron & Steel Business

Brittle crack arrest properties in ship construction have Electro gas arc welding
become more important as shipbuilding steel plates
become thicker and stronger. There have been indications CO2 arc welding
that steel toughness can have the effect of arresting Hatch side coaming
(Vertical member)
brittle crack initiating in a welded joint; however the Strength deck
T-weld joints of an actual large structure have not been (Horizontal member)
well investigated. This report describes how the brittle
crack length and steel toughness were found to affect
brittle crack arrest behavior. Furthermore, it was
suggested that brittle crack could be arrested by using Shift of joint

a horizontal plate of Kca that would be sufficient even

Fig. 1 Construction around hatch coaming
for the T-weld joint of an actual large structure.

Introduction (brittle crack arrestability), and if the positions of

welded joints are shifted away from each other in
The upsizing of container ships not only increases such a way that the crack collides directly with the
the allowable load and income per passage, but also base material of the deck. Several studies have been
decreases the number of crossings, which leads to the conducted on this possibility 3)- 6). Studies conducted
reduction of CO2 emissions and contributes to the by the authors indicate that brittle cracks can be
lessening of the environmental burden. The demand arrested by constructing the strength deck with a
for marine transport is increasing with the recent steel plate having a brittle crack arrestability (Kca)
rapid growth of East Asian economies such as that of exceeding a certain level 5), 6). These studies, however,
China. This upsizing trend is more apparent for have been conducted on small dimensional scales
transport vessels such as container ships 1). simulating the actual hull structure, leaving unclear
A container ship has an upper deck provided the Kca for steel plates that can arrest the long and
with a large opening and employs ultra-thick plates large cracks that may occur in actual hull structures.
for the longitudinal strength members of its hull. Hence, large specimens were prepared, each
As the ships are upsized, they require steel plates simulating a T joint between a hatch side coaming
that are stronger and thicker 2). According to an and strength deck. Brittle fracture tests were
experimental result 1), a brittle crack, initiated in a performed on these specimens while varying the
high-heat-input welded portion of a thick plate, specimen dimensions and the Kca of the plate
propagates straight along the welded portion. This simulating the strength decks. The performance of
poses a challenge when adopting thick plates and arresting long and large brittle cracks was studied,
joining them to hull structures by high-heat-input as reported in this paper.
welding. To satisfy the recent upsizing requirement
for container ships while assuring safety, brittle 1. Testing procedure
cracks must be arrested without fail before they
propagate along the high-heat-input welded portions 1.1 Specimen shape
of thick plates.
As shown in Fig. 1, a container ship has an upper Three types of specimens (specimens 1 - 3, as
deck (strength deck), a hatch side coaming and a T- shown in Fig. 2 (a)-(c)) were prepared by welding
weld joint that connects the hatch side coaming (a steel plates having a thickness of 60mm. Each
part of the longitudinal stiffening member) with specimen had a horizontal member simulating a
the upper deck. In this construction, a brittle crack strength deck and a vertical member simulating a
propagating straight along a welded portion of the hatch side coaming. Full penetration welding was
hatch side coaming can be arrested by the base performed to join the vertical member with the
material of the upper deck, if the base material has horizontal member to simulate a T joint between a
sufficiently high crack-arresting performance hatch side coaming and strength deck. Experiments

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 70

 Electro gas arc welding Table 1 Welding condition for specimen 1
400 400 800 Welding method Electro gas arc welding
Electro gas arc
Vertical Welding consumables DWS-1LG/φ1.6mm, CO2
welding
300

plate Welding conditions Heat input：35kJ/mm

Welding method CO2 arc welding
Welding consumables DW-55E/φ1.2mm, CO2
300

Horizontal Full penetration

t60 Welding current：190∼236A
CO2 welding plate welding
Welding conditions Welding voltage ：23∼29V
(a) Specimen 1 Welding speed：18∼60cm/min
Electro gas arc welding
400 400 700
Full penetration welding
Vertical
plate
Vertical Vertical plate
500

plate

Horizontal
plate
Horizontal plate
Horizontal
500

CO2 welding t60

plate Fig. 3 Cross section of T-weld joint of specimen 1

(b) Specimen 2 Table 2 Steels used for the specimen (Steel A was used for
vertical plate)
Electro gas arc welding
Horizontal plate
400 400 300 Dimension (Upper deck)
Specimen of 3/2
Specimen Kca at −10℃(N/mm )
(Material)
1 Fig.2(a) 4,200(N/mm )
3/2
700

Vertical
2−1 (Steel B)
plate Fig.2(b)
2−2 7,360(N/mm )
3/2

3 Fig.2(c) (Steel C)

t60 Horizontal 5,000 Impact load

CO2 welding plate
700

Specimen
1,200

Attached
horizontal plate
(c) Specimen 3
Attached plate Thermo couples
Fig. 2 Dimension of specimens (Unit：mm) Crack gauge for measuring crack propagation speed

Fig. 4 Schematic illustration of experiment for T-weld specimen

were conducted on these three types of specimens,

each having different dimensions, so as to study the for horizontal members to study the effect of the
effect of crack length reaching the horizontal plate. Kca values of the horizontal steel plates. The Kca
Each vertical member was made of two steel plates values at -10℃ of the steel plates B and C were
butt-welded by electrogas arc welding (EG welding). 4,200N/mm3/2 and 7,360N/mm3/2, respectively. Steel
The vertical members were welded to the horizontal plates B and C are both in the same strength class.
members by CO2 welding with full penetration. Table 2 shows the specimen numbers and the steel
Table 1 summarizes the welding conditions, and plates used for each specimen.
Fig. 3 shows a typical cross-sectional macro structure
of the full-penetration welded joint. 1.3 Testing method

1.2 Steel plates used for testing A tensile testing machine with a maximum load
of 30MN was used for the test. Fig. 4 illustrates the
The vertical member was made of the same steel testing method. A jig was disposed between the
(steel plate A) as reported previously (YS＝520MPa, testing machine and a specimen to apply homogeneous
TS＝619MPa)6). In this study, two types of steel plates stress to the specimen. The temperature of each
(Steel plates B and C), having different Kca, were used specimen was monitored using thermocouples

71 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

affixed to it at several locations. Each specimen the horizontal member did not break, it was ductile-
was homogenized at a set temperature of -10℃ for fractured forcefully by applying a tensile load after
more than 30 minutes before the test. The tensile the test.
load was controlled so as to make the average strain, In the case of Specimen 1 (vertical plate height
determined by strain gauges affixed to several (H)=300mm: horizontal plate, Steel plate B), the
location of the horizontal plate, become 257MPa (i.e., brittle crack propagated along the EG weld portion of
the design stress of a strength deck). the vertical plate and propagated further into the
Under the above conditions of temperature and horizontal plate via the T joint portion, as shown in
loading, an impact load was applied to the notched Fig. 5(a); however, the extent of the propagation
portion on the top of each specimen to initiate a was small compared with the thickness of plate B,
brittle crack. showing that the brittle crack was arrested
In a case where the propagation rate of brittle immediately after its propagation into steel plate B.
crack is not high enough, the load applied to the In the case of Specimen 2-1 (H＝500mm: horizontal
specimen may decrease after the initiation of the plate, Steel plate B), the brittle crack propagated to
brittle crack, which affects the test result. Therefore, the end of the horizontal member without being
an additional strain gauge affixed to the jig was used arrested (Fig. 5(b)).
for this test, so that the load behavior after brittle In the case of Specimen 2-2 (H＝500mm: horizontal
crack initiation could be monitored by measuring the plate, Steel plate C), the brittle crack propagated
strain on the jig. about 6~8mm into Steel plate C via the T joint and
was arrested (Fig. 5(c)).
2. Test results In the case of Specimen 3 (H＝700mm, horizontal
plate: Steel plate C), the brittle crack propagated
2.1 Brittle crack propagation behavior in vertical about 6~8mm into Steel plate C via the T joint and
member was arrested (Fig. 5 (d)).

In all of the above tests, brittle cracks propagated 3. Factors affecting brittle crack arrest characteristics
along the EG weld portions immediately after the
cracks were initiated. The crack propagation rate, 3.1 Discussion on the effect of Kca of steel plate for
immediately before the brittle cracks reached the T horizontal plates
joint, was measured to be 500-700m/s. The strain
gauges placed on the jig showed no load reduction Specimen 2-2 arrested the crack, unlike Specimen
that affected crack propagation. 2-1, despite the fact that they both have the same
shape (H＝500mm). This is attributable to the Kca of
2.2 Brittle crack arrest behavior of horizontal plate the horizontal steel plate of Specimen 2-2 being
higher than that of Specimen 2-1. The Kca of the
Fig. 5 (a)-(d) shows the fracture surfaces after horizontal steel plate significantly affects the brittle
the tests. It should be noted that, in the case where crack arrest characteristics at the T joint of the
horizontal plate.
propagate

propagate

3.2 Discussion on the effect of specimen dimensions

Crack

Crack

(brittle crack length)

3.2.1 Comparison of Specimens 1 and 2-1

Arrest

(a) Specimen 1 (b) Specimen 2 − 1

The brittle crack propagated in Specimen 1 (H＝
300mm) was arrested as soon as it collided with
the horizontal plate. On the other hand, the brittle
propagate

propagate

crack in Specimen 2-1 (H＝500mm) was not arrested,

Crack

Crack

although its horizontal plate had the same Kca as that

of Specimen 1. The difference is considered to be due
to the difference in the height H of the vertical plates.
Specimen 2-1 had a vertical plate with a larger H,
Arrest Arrest
which made the stress intensity factor at the crack tip,
(c) Specimen 2 − 2 (d) Specimen 3 K value, larger than that of Specimen 1 when the
Fig. 5 Test results (fracture surface) crack reaches the horizontal plate. The stress

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 72

 intensity factor exceeded the Kca of the horizontal 400 400
plate, causing the crack to propagate. E

3.2.2 Comparison of specimens 2-2 and 3; possibility

450
of arresting even longer and larger brittle crack
50
Vertical

700
plate
In the case of Specimen 3, the crack was arrested
despite the fact that the crack, when it reached the Strain gauge D

horizontal plate, was longer than the crack in

Specimen 2 - 2. This is considered to be caused by the
Horizontal
K value at the crack tip still being smaller than the Kca E plate
of the horizontal plate even under the conditions of
the crack in Specimen 3. Crack

700
propagation
According to Machida et al.7), a rapidly propagating
brittle crack suppresses the formation of a plastic
region, making the K value almost constant for the
increase of the crack length (a saturation phenomenon). Fig. 6 Schematic illustration of strain gauge measurement
If the K value of Specimen 3 is saturated, exhibiting along crack path
the same saturation phenomenon, there should be no
increase in the K value even for longer and larger Time when the crack passed
Crack start
cracks. If this is the case, cracks should be arrested by by the side of the strain gauge D
using Steel C for the horizontal member. The 3,500
behavior of the crack in this saturation phenomenon Strain value (μ strain)
3,000
in K value, however, has not been fully understood 2,500
because of the difficulty in evaluating the K value at 2,000

the tip of the brittle crack. This study attempts to 1,500

1,000
evaluate this saturation behavior in K value of a 0.042ms
500
brittle crack that propagates in a vertical member
0
during the T joint test. Measurements were 0 0.5 1
Time (ms)
conducted using a strain gauge to discuss the
Fig. 7 Strain behavior of strain gauge D in Fig.6
characteristics required for a horizontal member
arresting a long and large brittle crack.
higher diagonally in front, rather than right beside,
4. Steel plate characteristics required for arresting the crack in the vicinity of a crack tip. As shown in
long and large brittle cracks Fig. 7, the difference between the time when the
strain reached a maximum and the time when the
4.1 Study on the saturation behavior of K value crack passed right beside the strain gauge was about
0.042ms, which corresponds to a crack propagation
To study the saturation behavior of K value length of 26.7mm, considering the crack propagation
during brittle crack propagation, a strain gauge is rate (636m/s for an equivalent crack length)
affixed in the vicinity of the weld portion (i.e., the measured simultaneously. The distance between
crack propagation path) of Specimen 3 to evaluate the crack tip and the position where the strain
the strain when the crack passes (Fig. 6). The strain maximizes is calculated, according to the linear
distribution in the vicinity of a crack tip is considered fracture mechanics, to be 20mm in the crack
to vary corresponding with the K value at the crack propagation direction, assuming the positional
tip. So the behavior of the K value is discussed from relationship of this study (i.e., the strain measured
the aspect of strain behavior. at a point 50mm from the side of the crack path).
Fig. 7 shows an example of the values measured This calculated value is essentially the same as the
by the strain gauge. Also included in this figure is the measurement result, verifying the validity of the
time period (evaluated by a crack gauge measurement) evaluation, using strain measurement, of the strain
during which the brittle crack passed right beside the distribution in the vicinity of a crack tip.
strain gauge. This figure indicates that the strain Here, the maximum strain value was first
value reached a maximum immediately before the determined from the values measured by the strain
brittle crack passed right beside the strain gauge. gauges, then, the crack length which gives the
This is considered to be caused by strain that is maximum strain was estimated from the strain

73 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

propagation rate to establish the relationship small. To statically simulate this condition, in many
between the maximum strain and crack length. cases the original single-edge cracked specimen is
converted into a center-cracked specimen for elastic
4.2 Results of strain measurement calculation. In this study, a static elastic analysis was
conducted on the specimen converted into a center-
Fig. 8 shows the relationship between the cracked one with the symmetry boundary condition
maximum strain measured and the crack length. at the F portion in Fig. 9 (corresponding to the upper
Fig. 8 shows a tendency for the maximum strain end of the specimen). This calculation assumes that
measured to become almost constant when the a constant load is applied in the arrow direction at
crack length exceeds about 500mm. Also shown is a the end of the analysis model as shown in the figure,
phenomenon in which the strain drops rapidly at a such that the nominal stress becomes 257MPa. The
point right before (about 50mm before) the crack specimen length in the load direction was 5,000mm,
reaches the horizontal plate. which matches the distance between the loading
These measured results were compared with a points at the time of experiment illustrated in Fig 4.
finite element method (FEM) analysis performed on An analysis solver, ABAQUS6.5 8), was used.
Specimen 3. Fig. 9 shows the analysis model. Here, The maximum values for strain in the x direction
a 1/2 (half) structure in z-direction was modeled at each point / position corresponding to the location
using solid elements. The horizontal plate was where a strain gauge was affixed (on the line E-E in
divided into three elements, and the vertical plate Fig 6) were determined from the analysis result in
into six elements, in the plate thickness direction. order to examine their relationship with the crack
The adjacent area of the crack propagation was length. The results are shown by the dashed line in
segmented with meshes as small as 5mm at the Fig 8. As described previously, this study conducted
minimum, to improve the accuracy of the calculation. a static elastic analysis for a high speed deformation
A model crack (crack length, 0-700mm: crack tip, behavior, and the analysis result did not quantitatively
linear) was introduced, penetrating the plate thickness match the measured result. However, the analysis is
at the upper end of the vertical plate. considered to allow a qualitative comparison with
It should be noted that brittle cracks propagate the measured result. As shown in Fig. 8, the analysis
at high velocities (maximum 700m/s in Specimen 3), result shows a rapid decrease in strain immediately
making the materials through which the cracks before (about 50mm before) the crack reaches the
propagate deform very rapidly. As a result, the yield horizontal plate, the same decrease as measured,
strength of the material surrounding a crack tip indicating that the rigidity in front of the crack tip is
becomes very high due to the effect of strain rate increased by the horizontal plate. On the other hand,
dependence, and the crack opening becomes very no strain saturation phenomenon, as observed in
the actual experiment, exists in the simulated result
E
0
in the crack length range of 500mm or longer.
Measurement
100 FEM analysis Vertical
The strain distribution at a crack tip is significantly
Crack length (mm)

200 plate affected not only by crack length, but also by the
50
300 crack propagation rate. The strain saturation behavior
400 observed in this study might also have been affected
500 by the rapid crack propagation. To clarify the effect
600
of the propagation rate of a brittle crack, the relation
700 Horizontal plate Horizontal plate between the brittle crack propagation rate (measured
800 E
0 1,000 2,000 3,000 4,000 5,000 by a crack gauge) and crack length in Specimen 3 is
Maximum strain on E-E line
(μ strain) summarized in Fig.10. This figure indicates that a
Fig. 8 Relationship between crack length and maximum
brittle crack longer than about 200mm propagates
strain on E-E line in Fig.6 stably at a high rate of about 630-720m/s. In Fig. 8,
the measured strain value becomes almost constant
F Crack (length：0∼700mm) in the crack length range of about 500-620mm. In
Fig.10, no significant change in the crack propagation
rate is observed in the same crack length range,
indicating no significant change occurring in the
5,000 crack propagation behavior itself. In Fig.10, the
y y crack propagation rate increases by about 50m/s for
z x z x a crack length of 630mm, implying some effect on the
Fig. 9 FEM analysis model for specimen 3 periphery of the crack tip, however, no significant

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 74

 1,000 Conclusions
Crack propagation rate (m/s) 900
800
A study was conducted using large joint

Horizontal plate
700
600 specimens, each simulating a joint between a hatch
500
400
side coaming and strength deck. The focus was
300 put on the effect of the properties of a horizontal
200 plate (simulating a strength deck) on the arrest
100
characteristics of brittle cracks generated in the weld
0
0 100 200 300 400 500 600 700 800 portion of a vertical plate (simulating a hatch side
Crack length (mm)
coaming), taking into account the size of the
Fig.10 Relationship between brittle crack propagation rate specimen.
and crack length
The following describes the knowledge obtained:
・For the horizontal steel plate, Kca and specimen
increase in the crack propagation rate is observed for dimension (brittle crack length) significantly
the a crack length of 566mm. As shown in Fig. 8, affect the crack arrest performance. Steel C (Kca＝
strain saturation behavior is observed for a crack 7,360N/mm3/2) exhibits crack arrest even for the
length of 500mm-620mm. The change in the crack largest specimen used in this study.
propagation rate, observed in the same region in ・A strain measurement conducted during the
Fig.10, is not considered to affect the strain testing of Specimen 3 confirmed that the strain in
saturation behavior. the vicinity of a crack tip crack becomes saturated
during crack propagation once the crack reaches
4.3 Saturation behavior of K value and the a certain length.
characteristics required for steel ・The strain in the vicinity of crack tip is considered
to correlate with the K value. The brittle crack
From the previous results, it can be concluded propagation test conducted on Specimen 3
that the strain saturation phenomenon observed in indicates that the K value can possibly be
the actual measurements does not correlate with the saturated during crack propagation.
change in the brittle crack propagation rate and has ・From the above, cracks longer and larger than the
no link to brittle crack propagation behavior. Kinetic ones studied this time can possibly be arrested by
effect, which cannot be taken into account in FEM a horizontal plate (e.g., strength deck) made of a
static elastic analysis, is considered to have affected steel having adequate brittle crack arrest
the result. As described above, the strain distribution characteristics.
in the vicinity of a crack correlates well with the K
value of the crack tip, and thus the K of the crack tip References
in Specimen 3 may also have been saturated.
Assuming that this agrees with the above mentioned 1) Y. YAMAGUCHI, Kanrin, No.3, (2005), p.70.
comment by Machida, K values will saturate to the 2) S. OHKITA, Welding Technology, 2007-3, (2007), p.65
3) T. INOUE et al., Conference Proceedings JASNAOE, Vol.3
same value for longer and larger cracks, meaning (2006), p.359.
that brittle cracks can be arrested by using Steel plate 4) T. HANDA et al., Conference Proceedings JASNAOE, Vol.4
C for the horizontal plate (e.g., strength deck) as in (2007), p.459.
the case of Specimen 3. 5) E. TAMURA et al., Conference Proceedings JASNAOE, Vol.4
(2007), p.455.
6) E. TAMURA et al., R&D KOBE STEEL ENGINEERING
REPORTS, Vol.58, No.1(2008), p.8.
7) S. MACHIDA et al., Journal of the Society of Naval Architects
of Japan, Vol.131, (1972), p.367.
8) ABAQUS/Standard User's Manual, Hibbit, Karlsson & Sorensen,
Inc., 1999.

75 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

 Effects of Ca Addition on Formation Behavior of TiN
Particles and HAZ Toughness in Large-Heat-Input Welding
Taku KATO＊1, Shinsuke SATO＊2, Dr. Hiroki OHTA＊2, Toyoaki SHIWAKU＊3
＊1
Plate Products Development Department, Research & Development Laboratory, Iron & Steel Business
＊2
Ironmaking & Steelmaking Development Department, Research & Development Laboratory, Iron & Steel Business
＊3
Shinko Research Co., Ltd.

A new process technique, Kobe super toughness (KST), α2), 3). Kobe Steel has also confirmed that the addition
enables maintaining excellent toughness in heat affected of Ca improves HAZ toughness 4). However, not
zones (HAZs) formed by large heat input welding. In much detailed study has been done on the effect of
order to improve the KST treatment, the effect of Ca Ca addition on HAZ toughness and on TiN particles,
addition on HAZ toughness was studied with focus on leaving much unknown. Therefore, a detailed study
TiN particles. It was found that Ca addition inhibits the was conducted to ascertain how Ca addition affects
crystallization of coarse TiN particles and increases the particle size and the number of TiN particles. Based
number of fine TiN particles precipitated. Consequently, on the knowledge obtained, the amount of Ca to be
a fine-grained microstructure of HAZ was achieved and added was optimized for HAZ toughness and the
HAZ toughness was improved significantly. improvement effect was verified. The following is an
outline of this study.
Introduction
1. Experimental procedure
Heavier gauge steel plates are more and more
being used, with the recent increase in the size of Table 1 shows the chemical compositions of the
steel structures such as container vessels and steels tested in this study. Calcium mainly serves to
buildings. Welding such heavy gauge steel plates form oxide. The addition of Al, another oxide-
requires increased number of passes, making forming element, which is similar to Ca, was
conventional multi-pass welding highly inefficient. varied to vary the amount of dissolved oxygen
Therefore, ultra-high-heat-input welding, such as which affects oxide formation. Each composition was
electrogas arc welding and electroslag welding, melted in a 150kg vacuum melting furnace and was
which enable one-pass welding, is being increasingly cast in an ingot having a diameter of 250mm and a
used. The increased welding heat input, however, height of 400mm. Each ingot was heated to 1,100℃
keeps the heat-affected zones (HAZs) of the welds and hot rolled into a plate 50mm thick. The
at a high temperature for an extended period of temperature at the completion of the hot rolling was
time, as well as decreasing the cooling rate. This adjusted so as to be 850℃.
coarsens the microstructure and significantly Inclusions in each ingot were observed to study
decreases toughness. the effect of Ca addition on the crystallization of TiN.
To resolve such issues, Kobe Steel developed a The observations were conducted at intermediate
process technique, Kobe Super Toughness (KST).1) This positions between the ingot surface and the center.
technique enables the refinement of microstructure A field emission electron probe micro-analyzer (FE-
with a minor addition of Ti to disperse fine TiN EPMA) was used to observe the inclusions, to
particles, which suppress the coarsening of austenite analyze the composition of inclusions larger than
(γ) grains in HAZs and serve as transformation 2μm and to determine the particle size of TiN. Each
nuclei for intragranular ferrite (α). This ensures particle size of TiN was given by the diameter of a
favorable toughness in ultra-high-heat-input welded circle whose area is equal to the area occupied by
joints in HAZs, providing steel with a high degree of the TiN in each inclusion.
safety. To study the effect of Ca addition on the
The requirement for HAZ toughness, on the precipitation of TiN in a HAZ, a heat cycle test, called
other hand, is becoming increasingly stringent, a synthetic weld thermal cycle, which reproduces
which calls for further improvement of the KST and simulates ultra-high-heat-input welding, was
technique. There are reports that HAZ toughness conducted on rolled steels with the additive amount
can be improved by utilizing CaO and CaS, along of Al fixed at 0.03%. The heat cycle was applied to
with TiN, as transformation nuclei for intragranular specimens, each of which was taken from the mid-

Table 1 Chemical compositions of steels (mass％)

C Si Mn Al Ti Ca N others
0.08 0.35 1.55 0.013∼0.058 0.015 0∼0.0022 0.0060 Cu, Ni, B

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 76

 plane between the rolled surface and the center plane crystallized on these oxides. It should be noted
of a plate. To reproduce the welding conditions, that all the oxides in these compositions consist
these specimens were heated to 1,400℃, held at that essentially of CaO and Al2O3, or in other words, the
temperature for 30 seconds and then cooled in such remainder of CaO is almost entirely Al2O3. As a
a way that the temperature went from 800℃ to 500℃ whole, the particle size of TiN decreases with the
in 730 seconds. increase in CaO content. The size of TiN particles
After the synthetic weld thermal cycle, inclusions rapidly decreases in the CaO concentration range of
larger than 2μm were measured for their TiN about 40-60% and reaches a minimum at about 50%.
content by the FE-EPMA. The number density of TiN Fig. 3 shows the relationship between the CaS
precipitates smaller than 300nm was determined using content and the TiN content in the inclusions. The
a transmission electron microscope (TEM). TiN content decreases rapidly with the increase in
CaS content, and almost no TiN crystalizes out in the
2. Experimental results CaS content range higher than 20%.

2.1 Effect of Ca addition on crystallization of TiN 2.2 Effect of Ca addition on TiN precipitation in HAZ
in ingots
Fig. 4 shows the effect of Ca addition on the TiN
Fig. 1 shows typical inclusions observed by the content in the inclusions larger than 2μm and on
FE-EPMA. Steel with no addition of Ca exhibits the number density of fine TiN smaller than 300nm.
coarse TiN particles with sizes of around several With the increase in the amount of Ca added, the TiN
microns, which have been nucleated from Al2O3. The content in the inclusions larger than 2μm decreases,
steel with added Ca, on the other hand, exhibits while the number density of fine TiN increases. This
complex oxides consisting of Al and Ca, with CaS is considered to be caused by the increased addition
surrounding the complex. Almost no crystallization
of TiN was observed. 70
Fig. 2 shows the relation between the CaO
CaS content in inclusions (mass％)

60
content in the oxides and the size of TiN particles
50
Ca＝0ppm Ca＝14ppm
40
TiN
30
Observed
by FE-EPMA CaO−Al2O3 20
CaS
Al2O3 10
2μm 2μm
0
Al2O3 CaS 0 10 20 30 40 50
Frame
TiN content in inclusions (mass％)
format
TiN CaO−Al2O3 Fig. 3 Relationship between TiN content and CaS content in
5） inclusions
Fig. 1 Effect of Ca on inclusion morphology

25 70
Number of TiN particles below 300nm (10 2/μm2 )

8.0
TiN content in inclusions over 2μm (mass％)
Diameter of TiN particles on oxide (μm)

60
7.0
20
Envelope curve of 50
6.0
maximum value
5.0 15
40

4.0
30
10
3.0
20
2.0
5
1.0 10

0.0 0 0
0 20 40 60 80 0 5 10 15 20 25
CaO content in oxides (mass％) Ca content in steels (ppm)

Fig. 2 Effect of CaO content in oxides on diameter of TiN Fig. 4 Effect of Ca content in steels on coarse TiN particles
particles and fine TiN particles

77 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

of Ca suppressing the crystallization of coarse TiN considered to have caused the rapid decrease of TiN
on the oxides, resulting in a large portion of TiN content with the increase of CaS content, as shown
finely precipitated. in Fig. 3. Among the three compounds, CaO is the
most likely to become the nucleus for the
3. Discussion crystallization of TiN. However, as shown in Fig. 2,
the particle size of TiN tends to decrease as the CaO
As shown Fig. 1, TiN crystallizes out preferentially content increases. This is considered to be caused by
on Al2O3, but not on CaS or on CaO-Al2O3. In order the increased addition of Ca, which not only
to evaluate how easily TiN can crystallize out, the increases the CaO content, but also increases the
lattice misfits agsinst TiN and critical nucleus concentration of CaS, which is unlikely to become
formation energies of CaO, CaS and Al2O3 were the nucleus for the crystallization of TiN. This is
calculated according to the method proposed by considered to have caused CaS, having critical
Aaronson, J.H.Van Der Merwe, Jimbo and Morikage nucleus formation energy smaller for CaO than for
et al.6) - 8) The results are shown in Fig. 5 and Fig. 6. TiN, to be formed around the oxide as shown in
Here, the critical nucleus formation energy is Fig. 1, decreasing the area of contact between the
standardized by the nucleus formation energy of molten steel and CaO.
Al2O3 for comparison. Moreover, for the calculation, The following discusses the phenomenon, as
the interface energy was assumed to be the same as shown in Fig. 2, in which the particle size of TiN
the structural energy. The result indicates that CaS decreases rapidly in the CaO concentration range of
has the largest lattice misfit against TiN, while Al2O3 40-60%. In order for oxides to serve as the nuclei for
and CaO have almost the same level of lattice misfit. the crystallization of TiN, it must exist as a solid in
The critical nucleus formation energy was the largest molten steel. Taking this into account, our research
for CaS and the smallest for CaO. Therefore, from focused on the melting point of oxides. It is generally
the aspects of lattice misfit and critical nucleus known that, when the CaO content in an oxide
formation energy, CaS is least likely to become the reaches 40-60%, the melting point of the oxide
nucleus for the crystallization of TiN. This is becomes 1,413-1,600℃ and is minimized at the CaO
content of 49% 9). Assuming the temperature of
0.35
molten steel during casting to be 1,600℃, the region
where the content of CaO becomes 40-60% coincides
0.30 with the region where the melting point of the oxide
becomes lower than the temperature of the molten
0.25
steel. In other words, as the CaO concentration
Lattice misfit

0.20 approaches 49% in this region, the oxide in the

molten steel remains as a liquid at lower temperatures,
0.15
lowering the crystallization temperature of TiN and
0.10 shortening the time from the start of crystallization
until the completion of the solidification of molten
0.05
steel. This is considered to have suppressed the
0.00 coarsening of TiN.
Al2O3 CaO CaS
Fig. 5 Lattice misfit between TiN and inclusions
4. Verification of improvement in HAZ toughness
1.5
Normalized critical nucleus formation energy

1.4
Based on the above, the effect of Ca addition on
Compared to normalized TiN particles is summarized as follows. From the
1.3 energy of Al2O3 as 1.0
aspects of the melting point of oxides, lattice misfit
1.2 and critical nucleus formation energy, an increase in
1.1 Ca addition not only suppresses the formation of
1.0 coarse TiN, which becomes the origin of brittle
0.9
fracture, but increases the fine TiN which contributes
to the refinement of γ grain and the generation of
0.8
intragranular α. All of these are considered to be
0.7
effective in improving the HAZ toughness of ultra-
0.6 high heat-input welds.
Al2O3 CaO CaS
Fig. 6 Critical nucleus formation energy between TiN and
In order to verify this, the chemical compositions
inclusions shown in Table 1 were modified such that they

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 78

 have an additive amount of Al set at 0.03%. The Ca＝0ppm Ca＝21ppm
vTrs＝＋40℃ vTrs＝−34℃
steels were cast, rolled and subjected to the synthetic
weld thermal cycle, as described previously; then
Charpy tests were conducted to evaluate their HAZ
toughness. The fracture surfaces of the specimens
after the Charpy test were observed. Also observed
were micro structures of the specimens etched by
NITAL (an etching solution consisting of alcohol and 200μm 200μm

nitric acid).
Fig. 7 shows the relationship between the amount Fig. 9 Effect of Ca content on microstructure of HAZ
of Ca added to the steel and the transition temperature
at which fracture appears (vTrs). The increase in the
amount of Ca has been verified to lower the vTrs, Conclusions
significantly improving the HAZ toughness of ultra-
high heat input welds. The fracture surfaces of the Kobe Super Toughness (KST), a process technique
Charpy test specimens clearly show that coarse TiN for finely dispersing TiN particles, was further
particles in steel without the Ca addition originate improved by the addition of Ca. This technique is
brittle fracture as shown in Fig. 8. Furthermore, the applicable to a wide variety of steel plates used in
steel without Ca added exhibits coarse ferrite grains various fields including shipbuilding and construction,
larger than 100μm, as shown in Fig. 9, while the contributing not only to the improvement of welding
steel with Ca added exhibits a fine and homogeneous efficiency, but also to the improvement of the safety
dispersion of ferrite with a grain size of around of steel structures, and to the benefit of the entire
20μm, a result of the promotion of intragranular α society.
generation. These results verify that the addition of
Ca suppresses the formation of coarse TiN while References
increasing fine TiN, thus refining the microstructure
in the HAZs of ultra-high heat input welds and 1) Y. KASAMATSU et al., Tetsu-to-Hagane, Vol.65, No.8 (1979),
improving their toughness. p.1222.
2) S. KITAMURA et al., Tetsu-to-Hagane, Vol.73, No.6 (1987),
p.677.
60 3) S. SUZUKI et al., JFE Technical Report, No.5 (2004), p.19.
4) M. YAMAUCHI, R&D KOBE STEEL ENGINEERING
40 REPORTS, Vol.50, No.3 (2000), p.16.
5) T. KATO et al., Japan Welding Society Preprints of the National
20 Meeting of JWS, No.86 (2010), p.92.
6) Iron and Steel Institute of Japan, Use of Fine Inclusions in
vTrs (℃)

0 Microstructure Control of Steels, 1995, p.75.

7) Y. MORIKAGE et al., Tetsu-to-Hagane, Vol.84, No.7 (1998),
−20 p.510.
8) I. Jimbo et al., Trans of the ISS, 1989, p.43.
−40 9) Verein Deutsher Eisenhuttenleute, SLAG ATLAS 2nd Edition,
1995, p.39.
−60
0 10 20 30
Ca content in steels (ppm)
Fig. 7 Effect of Ca content on HAZ toughness

TiN

10μm

Fig. 8 Crystallized TiN at originating point of brittle fracture in

Ca free steel

79 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

 Characteristics of Highly Formable 590-980MPa Grade
Hot-dip Galvannealed Steel Sheets for Automobiles
Dr. Yuichi FUTAMURA＊1, Masaaki MIURA＊1, Michitaka TSUNEZAWA＊2
＊1
Sheet Products Development Department, Research & Development Laboratory, Iron & Steel Business
＊2
Sheet Products Development Department, Research & Development Laboratory, Iron & Steel Business (currently with Sheet Products Marketing
& Technical Service, Iron & Steel Business)

A series of 590-980MPa grade hot-dip galvannealed ductility under the design concepts of ① homogenized
steel sheets has been developed in order to improve the microstructure, ② inhibited carbide precipitation, ③
formability of automotive body frame parts, such as the maximum amount of ductile ferrite, and ④ ferrite
lower pillars. These are either DP or TRIP-aided steel with its own strength increased by solid solution
sheets, designed 1) to have homogeneous microstructures, strengthening to prevent the deterioration of local
2) to inhibit the precipitation of carbide, 3) to maintain deformability. In addition, ⑤ the TRIP-aided steel
ferrite as much as possible, 4) to harden the ferrite to sheet is designed with its microstructure controlled
prevent degradation of local deformability, and in TRIP- so as to secure a large amount of retained austenite.
aided steel sheet, additionally, 5) to obtain a large As a means for achieving the above objectives, the
amount of retained austenite. The above microstructural DP steel sheets adopt the technology developed by
controls are mainly accomplished by the addition of Kobe Steel for 590 - 980MPa grade GA steel sheets 5) - 12)
silicon, which can be adopted through a special surface and have compositions that largely suppress the
preparation technology. The developed steel sheets have bainite transformation during the cooling after
not only excellent formability, but also meet practical annealing (②, ③). Furthermore, both the DP steel
performance requirements, including good spot sheet and TRIP-aided steel sheet developed this time
weldability and high coating. contain silicon (Si), an additive element playing an
important role in stabilizing the ferrite and
Introduction significant solid solution strengthening (③, ④). For
some time, Si has been known to be effective in
In recent years, high-strength steel sheets are improving workability 13); however, its application to
more and more being used for automotive structural GA steel sheet was difficult because the element
members to make auto bodies strong enough to meet tends to deteriorate coating quality. To resolve this
increasingly stringent collision safety standards issue, Kobe Steel developed a special surface
and light enough to meet the emission reduction modification which enabled GA steel sheet to contain
requirements for environmental protection 1). Among above 1.0% of Si, as in the case of cold-rolled steel
these sheets, galvannealed (hereinafter referred to as sheet, leading to this development.
"GA") steel sheets are applied to parts that require The addition of Si can significantly increase
corrosion resistance. In particular, high-strength elongation. On the other hand, Si addition facilitates
GA steel sheets with excellent workability, are the precipitation of ferrite in the microstructure
proactively adopted for lower parts of pillars and during cooling, which decreases the yield ratio,
other structural members. With the recent increase leading to a significant reduction of yield strength
in the number of parts employing high-strength steel for a given tensile strength. Thus, when producing
sheets, the demand for further improved workability a steel sheet containing Si, the microstructure is
is growing 2), 3). controlled while focusing on the recovery and
In response to these needs, Kobe Steel has recrystallization behaviors of the cold-rolled structure,
developed a high tensile-strength GA steel, in the in which the cold rolling reduction is optimized for
strength grade of 590 - 980MPa, which has an each composition and for sheet thickness by
elongation 1.3 times higher than that of conventional adjusting the yield strength so as not to cause the
materials 4). This paper describes the concept of its deterioration of elongation during annealing. This
microstructure control and major characteristics. control method offers a technique that is effective
in obtaining a homogeneous composite structure
1. Concept of microstructure control regardless of sheet thickness (①). TRIP-aided steel
sheets, on the other hand, adopt a microstructure
The developed materials include a dual-phase control which maximizes the amount of retained
(DP) steel sheet, consisting of ferrite and martensite, austenite. Typically, a TRIP-aided steel sheet is
and a transformation-induced plasticity (TRIP) aided austempered to increase carbon concentration in
steel sheet, consisting of ferrite, bainite and retained untransformed austenite. When producing a GA
austenite. Both kinds of steel sheets ensure high steel sheet, the steel sheet is exposed to an elevated

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 80

 temperature in galvannealing treatment, which electron backscattering pattern (EBSP) measurement.
induces a phenomenon, unique to GA steel sheet, In the grain boundary distribution diagram (Fig. 2
involving a structural change in which the austenite (b)), the black solid lines represent large angle grain
decomposes (precipitating carbide). The newly boundaries with crystal misorientation of 15 degrees
developed steel sheet contains retained austenite, or larger, while the gray solid lines represent small
whose amount is maximized by an optimized angle grain boundaries with crystal misorientation
austempering (⑤). This new austempering includes smaller than 15 degrees. The figure indicates that the
galvannealing, unlike conventional austempering ferrite structure contains many small angle grain
which simply condenses carbon. boundaries, which are remainders of a cold-rolled
Fig. 1 shows typical microstructures of the newly structure that has been annealed and is subgrained.
developed steel sheets: (a) DP steel sheet of 980MPa The newly developed materials make use of such
grade and (b) TRIP-aided steel sheet of 780MPa subgrain structures to adjust strength characteristics
grade. Both the steel sheets contain large amounts so as to achieve the high elongation described later.
of ferrite with no carbide precipitate observed. Fig. 3 shows cross-sectional SEM micrographs of
Furthermore, the DP steel sheet has a homogeneous the newly developed 590MPa grade and 980MPa
and fine composite structure, while the TRIP- grade DP steel sheets, both having the quality of
aided steel sheet is predominantly massive retained exhibiting homogeneous coating layers with
austenite with needle-like retained austenite favorable powdering resistance.
(indicated by the arrows in the figure) between the
laths of bainitic ferrite. Fig. 2 shows the grain boundary 2. Main characteristics of newly developed materials
distribution inside the 980MPa grade DP steel sheet
obtained by a crystal orientation analysis based on an 2.1 Formability

Table 1 summarizes the typical tensile properties

(a) 980MPa grade DP steel sheet (b) 780MPa grade TRIP steel sheet
of the newly developed materials (thickness 1.6mm).
Also included in this table are the typical values of
590 - 980MPa grade cold-rolled steel sheets, in
5μm
which the reference values of the 980MPa grade
cold-rolled steel sheet are represented by those of
Fig. 1 SEM images of microstructure in developed steel sheets a conventional Kobe Steel DP steel sheet that was
((a)980MPa grade DP and (b)780MPa grade TRIP) developed for higher elongation 14). All the newly
developed materials exhibit elongations equivalent
(a) IQ map (b) Boundary distribution to those of the reference cold-rolled steel sheet in
the same strength grade. Fig. 4 compares the tensile
strengths (TSs) and elongations (ELs) of the newly
developed sheets of 980MPa grade DP steel and
5μm 780MPa grade TRIP-aided steel with those of the
conventional materials. As the result of the
ND
composition design and microstructure control
RD described previously, both of the newly developed
High angle boundary sheets realize elongations up to 1.3 times greater
(more than 15 agree)
Low angle boundary 5μm than that of Kobe Steel's conventional materials. A
(below 15 agree)

Fig. 2 Image Quality map (a) and Crystal orientation image Table 1 Tensile properties of developed GA (Galvannealed)
map with boundary distribution (b) in developed steel sheets and reference CR (Cold-Rolled) steel
980MPa grade DP steel sheet sheets
YP TS EL
Steel TS grade Category
(MPa) (MPa) (％)
(a) 590MPa grade DP steel sheet (b) 980MPa grade DP steel sheet
590MPa Dual Phase 387 613 34
Galvannealed (GA) Dual Phase 481 828 23
steel 780MPa
(Developed) TRIP 478 823 29
980MPa Dual Phase 619 1,037 18
590MPa Dual Phase 388 633 33
5μm Cold-rolled (CR)
780MPa Dual Phase 509 838 22
steel
980MPa Dual Phase 635 1,032 18
Fig. 3 SEM images of coating layer in developed steel sheets ・Specimen thickness : 1.6mm
((a)590MPa grade, (b)980MPa grade DP steel) ・Tensile test : JIS Z2241 (JIS Z2201 ＃5 specimen in Transverse direction)

81 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

 30.0 Developed Table 2 Hole expansion ratio and Minimum bending radius
of developed and conventional steel sheets
TRIP
Minimum
25.0 Thickness λ-value
Steel TS grade Category bending radius
Conventional (mm) （%）
Elongation (％)

25,000 〈Transverse direction〉

20,000 590MPa Dual Phase 1.6 74 0.0
20.0
Developed Dual Phase 1.6 27 0.0
Developed 780MPa
15,000 steel TRIP 1.4 29 0.0
Dual-Phase 980MPa Dual Phase 1.6 27 0.0
15.0
590MPa Dual Phase 1.6 68 0.0
Conventional Conventional
TS×EL＝10,000 780MPa Dual Phase 1.6 31 0.5
steel
10.0 980MPa Dual Phase 1.6 25 2.0
700 800 900 1,000 1,100 1,200
・Stretch flanging (Hole expanding) test :
Tensile strength (MPa)
 Hole expansion ratio :λ-value obtained by method of JFST1001
Fig. 4 Relationship between tensile strength and elongation ・Bending test : V-block method (90 degree angle) according to JIS Z2248
in developed 980MPa grade DP and 780MPa grade  (JIS Z2203 ＃3 specimen in transverse direction)
TRIP steel sheets
0.6 (a) 590MPa grade DP steel sheet

stretch forming test confirmed that the TRIP-aided 0.4

Conventional
steel sheet exhibits extremely high strain dispersibility 590MPa grade steel sheet
0.2
with less likelihood of local thickness reduction and

εy
achieves a limit-forming height comparable with that 0.0
0.0 0.2 0.4 0.6 0.8
of a conventional 590MPa grade DP steel sheet which
εx
is one grade lower in terms of strength. −0.2

Table 2 shows the typical values of the hole

−0.4
expansion ratio (λvalue), an index for stretch-
flangeability, and the minimum bending radii in a 0.5 (b) 980MPa grade DP steel sheet
90 degree V-bending test. The minimum bending 0.4
radius was determined from the smallest bending 0.3
radius that causes no crack in a tested material bent by 0.2
a 90 degree punch with a tip radius (R) of 0 - 5.0mm. Conventional
0.1 980MPa grade steel sheet
εy

The direction of bending was vertical to the direction

0.0
of rolling. Also included in the table are the typical 0.0 0.1 0.2 0.3 0.4 0.5
−0.1
values for conventional 590 - 980MPa grade GA steel εx
−0.2
sheets. The newly developed materials have λ values
−0.3
that are all equivalent to those of the conventional
materials in the same strength grade, with Fig. 5 Forming limit diagrams of developed steel sheets
suppressed deterioration of local deformability, and ((a)590MPa, (b)980MPa grade DP)
as a result have superior balances of EL and λ
compared with conventional materials. In general, circle diameter is 0.25inch, and the dashed lines
bendability is known to correlate with stretch- represent the forming limits for the conventional
flangeability as well as with local ductility 15), 16). Thus materials in the same strength grade for comparison.
the minimum bending radius is regarded to correlate It has been shown that both of the newly developed
favorably with λ. The newly developed material, materials have forming limits higher than those for
however, exhibits an excellent bendability despite its conventional materials in a plane strain state, which
λ comparable to that of the conventional materials. is the most stringent forming condition. The high
This indicates that the bendability is not necessarily forming limit is attributable to the previously
determined by λ. In other words, there may be described composition design and microstructure
effects other than the generally-known one of the control, both having an effect of improving
addition of Si decreasing the difference in hardness elongation.
between ferrite and martensite, improving the local
deformability. However, the detailed mechanism has 2.2 Spot weldability
not been clarified yet, and a study is being conducted
to elucidate the details. Fig. 5 shows the forming Fig. 6 (a) and (b) respectively show the tensile
limits for (a) the newly developed material of shear strength (TSS) and cross tensile strength (CTS)
590MPa grade and (b) 980MPa grade DP steel sheet of the 590MPa grade and 980MPa grade DP steel
(both having a thickness of 1.2mm). The scribed sheets, both newly developed, for varying welding

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 82

 50,000 980MPa grade Table 3 Spot welding conditions for Fig.6
590MPa grade
Expulsion (980) Electrode tip Dome type Cu-Cr
Tensile shear strength (N)
40,000
Tip diameter 8mm

30,000 Electrode force 4,950N

Welding time 28cycle (60Hz)
20,000 Welding current 6-14kA
Expulsion (590)
Cooling water (Upper, Lower) 1.5L/min
10,000
JIS Z3140 A-class
ND：5.4mm min. load：13.5kN
0 Thickness 1.2mm 1.6mm 2.0mm
4 6 8 10 12 14
Tip diameter 6mm 8 mm 8mm
Welding current (kA)
Electrode force 3,750N 4,950N 6,450N
(a) Tensile shear strength
Welding time (60Hz) 23 cycle 28 cycle 29 cycle
30,000 980MPa grade 30,000
590MPa grade Expulsion (980)

TSS on ND：5.4mm (N)

Cross tensile strength (N)

25,000 25,000

20,000 20,000

15,000 15,000

10,000 10,000 JIS Z3140 A-class

min. load：13.5kN
5,000 Expulsion (590)
5,000
ND：5.4mm 1.2t 1.6t 2.0t
0 Thickness (mm)
4 6 8 10 12 14
Fig. 7 Relationship between tensile shear strength and
Welding current (kA)
(b) Cross tensile strength sheet thickness in developed 980MPa grade DP steel
sheet
Fig. 6 Relationship between welding current and (a)tensile
shear strength, (b)cross tensile strength in developed
Table 4 Spot weldability in developed 590MPa and 980MPa
590MPa and 980MPa grade DP steel sheets
grade DP steel sheets welded with/without gap
(1mm) between sheets
Gap Welding current
current. Each sample sheet has a thickness of 1.6mm TS grade
between A-class TSS :
welded under the conditions shown in Table 3. (steel) ND:5.4mm Expulsion
sheets 13.5kN
Both the 590MPa grade and 980MPa grade steel 590MPa 0mm 6.5kA 6.5kA 9.0kA
sheets newly developed, have tensile shear strengths (Developed steel) 1mm 6.5kA 6.5kA 9.0kA
exceeding the requirement of JIS-A grade; i.e., 13.5kN, 980MPa 0mm 6.0kA 5.5kA 8.0kA
(Developed steel) 1mm 6.0kA 5.5kA 8.0kA
with a nugget diameter of 5.4mm. A suitable current
range, not causing expulsion, is confirmed to be as
large as 2.5kA or wider. On the other hand, cross Table 4 summarizes the results of tests aimed at
tensile strength may decrease as an effect of the evaluating the practical spot weldability, in which
composition 17). However, the newly developed the newly developed 590MPa grade and 980MPa
materials exhibit no noticeable deterioration in grade materials were welded under the conditions
strength. The newly developed materials of 590MPa of a smaller electrode tip diameter (6mm) and lower
grade and 980MPa grade result in a ductility ratio, applied pressure (3,430N), compared with the
the ratio between the tensile shear strength and conditions shown in Table 3. In addition, the welding
cross tensile strength (CTS/TSS), of 0.65 or higher was performed with or without a gap (1mm) in
and 0.48 or higher, respectively, in the suitable between the welded sheets. The table also includes
welding current range. Fig. 7 shows the relationship the current values for the specified nugget diameter
between the tensile shear strength and sheet (5.4mm), JIS-A grade specification (13.5kN) and for
thickness for a nugget of the newly developed expulsion. For all the cases, the electric current range
980MPa grade material that has a diameter of in which both the specified nugget diameter and
5.4mm. For the thicker sheets, the electrode diameter JIS-A grade specifications are satisfied without
and applied pressure were increased and the current causing expulsion is above 1.5kA. This indicates that
period was extended. As a result, a favorable joint the newly developed materials not only have
strength with a tensile shear strength greatly favorable workability, but also produce high
exceeding the load specified by JIS-A class was welding joint strength in a consistent manner.
obtained.

83 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

Conclusions References

New galvannealed steel sheets of 590 - 980MPa 1) Y. KURIYAMA et al., Journal of Society of Automotive
grade with elongations as high as 1.3 times those Engineering of Japan, Vol.55, No.4 (2001), p.51.
2) S. KOBUKI, TOYOTA Technical Review, Vol.52, No.1 (2002),
of the conventional materials have been developed. p.8.
The concepts of their microstructure control and 3) K. SHIBATA, NISSAN Technical Review, No.50 (2002), p.26.
main properties have been introduced. The newly 4) Y. FUTAMURA et al., R&D Kobe Steel Engineering Reports,
developed materials are characterized not only by Vol.57, No.1 (2007), p.109.
their excellent elongations, but also by superior 5) M. NAKAYA et al., R&D Kobe Steel Engineering Reports,
Vol.50, No.1 (2000), p.75.
bendability, spot weldability and coating quality, 6) Y. OMIYA et al., R&D Kobe Steel Engineering Reports, Vol.52,
which are sufficient to satisfy customer needs. Kobe No.3 (2002), p.10.
Steel regards these newly developed materials as a 7) M. KAMURA et al., R&D Kobe Steel Engineering Reports,
high-elongation type among the high-strength GA Vol.51, No.2 (2001), p.79.
steel sheets. In response to customer needs, Kobe 8) M. KAMURA et al., IBEC2002, Proceedings of the 2002 IBEC
and ATT Conferences on CD-ROM, (2002), 2001-01-3094.
Steel has a lineup of products including, in addition 9) M. KAMURA et al., SAE Technical Paper (2003), 2003-01-0522.
to the above, steels with improved weldability and 10) X. M. CHEN et al., SAE Technical Paper (2004), 2004-01-1048.
steels of high YS type. On the other hand, recent 11) X. M. CHEN et al., SAE Technical Paper (2005), 2005-01-0354.
customer requirements for material properties have 12) Y. FUTAMURA et al., R&D Kobe Steel Engineering Reports,
become much more stringent than they were several Vol.57, No.2 (2007), p.11.
13) M. MIYAHARA et al., R&D Kobe Steel Engineering Reports,
years ago, so Kobe Steel will continue striving to Vol.35, No.4 (1985), p.92.
improve the characteristics of materials to resolve 14) T. TAMURA et al., R&D Kobe Steel Engineering Reports,
issues that the customers may have and thus expand Vol.52, No.3 (2002), p.6.
the application of high-strength steel sheets. 15) Y. TANAKA et al., R&D Kobe Steel Engineering Reports,
Vol.42, No.1 (1992), p.20.
16) J. IWAYA et al., JOURNAL OF THE JAPAN SOCIETY FOR
TECHNOLOGY OF PLASTICITY SOSEI-TO-KAKO (Journal
of JSTP), Vol.35, No.404 (1994), p.1122.
17) S. NOMURA et al., Technical Commission on Resistance
Welding (1981), RW-192-81.

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 84

 Formability of TRIP Type Banitic Ferrite Steel Sheet
Takayuki KIMURA＊1
＊1
Process Engineering Development Department, Research & Development Laboratory, Iron & Steel Business

In recent years, cold rolled steel sheets of 980MPa needed as the applications of high tensile strength
grade have been increasingly used for automotive steel of 980MPa grade increase. In response, Kobe
parts to improve collision safety (crashworthiness) and Steel aims at improving elongation characteristics,
to reduce body weight. Kobe Steel has developed a while maintaining stretch-flangeability in comparison
new 980MPa cold rolled steel sheet with elongation with the conventional 980MPa grade DP steel sheet of
properties that are an improvement over conventional a type focusing on elongation (hereinafter referred
dual-phase (DP) steel sheets. This article focuses on the to as "980DP steel"). Fig. 1 3) shows micrographs of
press formability of the newly developed steel sheet. the developed steel and 980DP steel. The developed
Press formability testing was performed using a small- steel contains a large amount of retained γ, which
sized model die and a large-sized actual part die. The is morphologically controlled so as to be finely
result clearly indicates that the developed steel sheet dispersed in elongated shapes to ensure elongation
has a significantly improved press formability when characteristics. The purpose of this is to increase the
compared with conventional DP steel sheets. stability of retained γ such that the TRIP effect
continues until the late stage of deformation.
Introduction
1.2 Tensile properties
The latest automobiles are required to have both
improved collision safety and weight reduction To evaluate the mechanical properties of the
and make wide use of high-strength steel sheets. developed steel, tensile tests were conducted on
However, steel sheets with higher strength tend to specimens prepared according to JIS 5, using a
have poor elongation characteristics and are more 100kN autograph manufactured by Shimadzu
prone to crack during press forming. Therefore, some Corporation. The cross-head speed was kept
auto parts must be made in separate pieces for ease of constant at 10mm/min. Comparisons were made
forming. To resolve the issue, several methods have with 980DP steel, as well as with 780MPa grade
been proposed, including multiple process steps DP steel (hereinafter referred to as "780DP steel").
for improving deep-drawing formability 1) and dual As shown in Table 1, the developed steel exhibits an
punching to improve stretch flangeability 2). So far elongation comparable to that of the 780DP steel
as the materials are concerned, the steel sheets that and greatly exceeds that of the 980DP steel, implying
have been developed are of high strength and its excellent formability. In addition, the developed
superior formability. Kobe Steel has developed a
cold-rolled, transformation-induced-plasticity (TRIP)
type, banitic-ferrite steel sheet of 980MPa grade a) Developed steel b) DP steel
(hereinafter referred to as the "developed steel"),
which is suitable for auto body frames, with
elongation characteristics that are an improvement
over those of conventional dual phase (hereinafter 5μm 5μm
referred to as "DP") steel sheets 3), 4).
This paper describes the features of the developed Fig. 1 Microstructure of 980MPa grade cold rolled steel sheets
a) developed steel, b) conventional DP steel 3）
steel and reports the results of formability tests
conducted using a small laboratory scale die and a
large die simulating actual parts. Table 1 Mechanical properties of sample steels
(t：1.4mm)
YS TS EL n value
1. Microstructure and mechanical properties of (MPa) (MPa) (％) 2-6％
developed material
a) 980 Developed steel 631 1,062 20 0.22

1.1 Microstructure b) 980 DP steel 642 1,060 16 0.15

c) 780 DP steel 527 831 20 0.13

Steel sheets with improved workability have been

85 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

steel exhibits a n value (work hardening index) 2. Formability of developed steel
which is much higher than those of the reference
steels, an indication of superior strain dispersibility. The developed steel sheet was studied for its
Thus, the developed steel advantageously suppresses formability in terms of the four major forming modes
the local reduction of thickness during press forming of stretching, deep drawing, stretch flanging and
and is expected to enhance collision performance bending. The thickness of the tested materials was
when used for parts in automobile body frames. 1.4mm.
In order to verify the superior strain dispersibility
of the developed steel, strain measurement using 2.1 Stretch formability
a non-contact strain measuring machine, ARGUS,
manufactured by GOM mbH, was conducted in the Fig. 4 depicts the testing apparatus. The stretch
longitudinal direction of the tensile specimens. A formability was evaluated by the maximum forming
pattern of equally-spaced dots was provided on each height, which is determined by the punch stroke,
specimen prior to the test. Each specimen was which causes failure with a rapid load drop in the
mounted on a tensile machine, elongated, and load-stroke diagram. As a reference, the 980DP steel
dismounted as soon as the strain reached 15%. After was also tested. The testing apparatus was a 500kN
the deformation, the position relationships in the universal deep-drawing tester manufactured by
dot pattern were measured and the image was TAKES-GROUP Ltd.
processed to evaluate the strain distribution. The The maximum forming heights shown in Fig. 5
measurement results, shown in Fig. 2, indicate that verify that the developed steel has a formability
the developed steel exhibits less local strain superior to that of the 980DP steel. This result is as
concentration when compared with the reference expected from the results of the tensile tests. The
steels, with the strain almost uniformly distributed stretchability is reported to be affected by the
within the specimen. The distributions in the elongation and n values of materials 5).
thickness reduction rates of the specimens (Fig. 3) As in the case of the tensile test, the strain
verify that local strain concentration is suppressed dispersion effect in the stretch formed samples was
for the developed steel. studied. Thickness reductions were measured on the
developed steel and 980DP steel, both formed to a
height of 17mm. As shown in Fig. 6, the thickness
30mm
reduction rate for the developed steel is less than
980 30% even near the top where the reduction reaches
Developed the maximum. On the other hand, the 980DP steel
steel
exhibits a thickness reduction of about 35% for the
same forming height, verifying the superior strain
980DP
steel dispersibility of the developed steel in the stretching
test.
780DP
steel
Punch：φ50mm, rp＝25mm rd
Die：φ54.8mm, rd＝10mm
(％) 22 20 18 16 14 12 10 8 6 4 2 0 Lubricant：NOX-RUST 530
Blank holding force：118kN
Fig. 2 Longitudinal strain distribution of tensile test specimen Blank size：150×150mm
(strain：15％) rp

0 Fig. 4 Experimental apparatus for stretch formability

Thickness reduction (％)

−2
Maximum forming height (mm)

20
0 measurement position
2
4 19
6
8 18
980 Developed steel
10
12 980DP steel 17
14 780DP steel
16 16
−80 −60 −40 −20 0 20 40 60 80
Measurement position (mm)
15
Fig. 3 Distribution of thickness reduction rate in tensile test 980 Developed steel 980DP steel
specimen (strain：15％) Fig. 5 Maximum forming height

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 86

 ：980 Developed steel
0
Thickness reduction (％)
：980DP steel
5 Punch：φ50mm, rp＝25mm
Measuring rd
10 Die：φ54.8mm, rd＝10mm
15 position
Lubricant：NOX-RUST 530
20 1
Blank holding force：9.8kN
25 Blank size：φ100mm
30 19
    (Drawing ratio＝2.0)
35 rp
40
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Position NO. Fig. 7 Experimental apparatus for drawing formability
Fig. 6 Thickness reduction rate distribution of stretch test
specimen (Forming height：17mm) ◆：980 Developed steel
▲：980DP steel
30 ●：780DP steel

Maximum forming height (mm)

◆
2.2 Deep drawing formability n value：0.22
980 Developed steel
28

Deep drawing tests were conducted using a 26

punch having a spherical head with a head diameter ●
of 50mm (Fig. 7). The maximum forming height at 24 n value：0.13
980DP steel
a drawing ratio of 2.0 was used as the evaluation ▲
index. The maximum forming height was determined 22 n value：0.15
by the punch stroke, which causes failure with a
rapid load drop in the load-stroke diagram, just as 20
17 18 19 20
780DP steel
in the case of the stretch formability test. Both the Total elongation (％)

980DP steel and 780DP steel were tested as reference Fig. 8 Maximum forming Fig. 9 Drawing test specimen
materials. The tests were conducted using a 500kN   height (Draw)
universal deep-drawing tester manufactured by
TAKES-GROUP. Ltd. steel, decreases the flow resistance and contributes to
As shown in Fig. 8, the developed steel exhibits the improved deep drawability of the developed
the best formability among the three types of steel steel.
tested. The difference can easily be seen in the photos
of the formed samples (Fig. 9). The developed steel 2.3 Stretch flangeability
exhibits the highest total elongation and n value, the
combined effect of which is considered to have led The stretch flangeability was evaluated by a hole-
to its superior formability. As described above, the expanding test. The test was conducted according to
developed steel contains retained γ, which is the Japan Iron and Steel Federation Standard,
considered to assure excellent strain dispersibility JFST1001. The stretch flangeability of the developed
and to increase the strength at the punch shoulder steel is almost equal to that of the 980DP steel, as
portion by strain-induced transformation as well 6). shown in Fig.10. In general, the stretchability and
In other words, the excellent strain dispersibility is stretch-flangeability of a high-strength steel sheet are
considered to allow the developed steel to have a in a trade-off relationship. In other words, steel sheet
larger thickness near the vertex in comparison with having a superior stretchability tends to have poor
the DP steel. In addition, the greater strength stretch-flangeability and vice versa. The reason for
increased by work hardening is considered to have the suppressed degradation of stretch-flangeability,
enabled the developed steel to endure great resistance despite the superior stretchability, compared with
against the material flow into the flange. Another the 980DP steel, is that the developed steel has a fine
reason for the excellent formability is that, unlike and homogeneous microstructure that suppresses
conventional retained γ steel, the developed steel local cracking.
has finely dispersed retained γ in elongated shapes,
such that work hardening continues until the late 2.4 Bending formability
stage of the deformation. As for resistance against the
material flow into the flange, the volume expansion Bending formability is evaluated by the
associated with the strain-induced transformation of existence/nonexistence of a crack on the outer
retained γ suppresses the transformation under the surface of a sample bent by V bending with a punch
compressive stress generated by the drawn material. with a punch angle of 90 degrees. The punch tip
Furthermore, the strength increment due to work radius, R, was varied from large to small. In this V
hardening, which is smaller than in the case of DP bending test, each sample was placed such that the

87 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

 evaluation was conducted based on an A-pillar,
70
which contains the deformation element of stretch
Limited hole expanding ratio (％) 60 flanging. A simulation die was used for evaluating
the overall formability of the entire part.
50 Fig.11 is a photo showing the formed part. The
40
A-pillars, currently in volume production, use steel
sheets of either 440MPa grade or 590MPa grade steel
30 sheets, and only a few employ the 980MPa grade. In
many cases, this component is constructed in
20
separate blocks, in which the pocket shape in the
10 lower part of the vehicle is separated to avoid
cracking during the press forming. The evaluation
0
980 980DP 780DP this time, on the other hand, was conducted by
Developed steel steel forming the part as a single piece without splitting.
steel
As shown in Fig.11, the press-formed shape requires
Fig.10 Limited hole expanding ratio of steels
a superb deep-drawability and stretchability, as
well as a superior stretch-flangeability at the blank
Table 2 Bendability of steels
holder portion. As a reference, 980DP steel was also
90°V press-formed. The sheet thickness in either case was
R
R＝0mm R＝0.5mm R＝1.0mm 1.4mm.
980 Developed steel △○ ○○ ○○ The blank holder force was varied from 800kN
980DP steel △△ ○○ ○○ to 2,000kN, while checking for the existence/
○：Good △：Hair crack nonexistence of cracks and wrinkles. As shown in
Table 3, the 980DP steel could be formed up to
rolling direction was parallel to the bending ridge 1,300kN without cracking; however, wrinkling
line, and the punch was forced until the applied occurred in the seat surface facing the upper surface
pressure reached 19.6kN. The test results are of the punch. These wrinkles, occurring inside the
summarized in Table 2. They confirm that the part, can be suppressed by increasing the blank
developed steel has a bending formability almost holder force. The arrow in the table indicates the
equivalent to that of the 980DP steel. The developed direction along which the wrinkle suppression
steel sheet can be bent to the point where it is
substantially U-shaped, which is considered to
satisfy the bending workability required for body
frame parts.

2.5 Formability evaluation using die simulating

front pillar

As described above, the developed steel is

superior in formability to the conventional type of
Fig.11 Shape of press test part
DP steel sheet in all three of the aspects of
stretchability, deep-drawability and bendability. Its Table 3 Results of press test
stretch-flangeability has been confirmed as almost
BHF fracture
equal to that of the 980DP steel. To demonstrate the (blank holding wrinkle
980 Developed 980DP
superior formability of the developed steel, an force) steel steel
evaluation was conducted using a large-sized die 800kN ○ ○ NG
simulating an actual part. 1,000kN ○ ○
Kobe Steel once evaluated the formability of the
1,200kN ○ ○
developed steel using a die to simulate a B-pillar and
1,300kN ○ ○
confirmed the superiority of the developed steel 3).
1,400kN ○ ×
The evaluation at that time, however, focused on the
formability of deformation elements including only 1,500kN ○ ×

stretching and deep drawing, rather than the 1,700kN ○ ×

formability of an entire part, lacking the deformation 2,000kN ○ × Good
element of stretch flanging; so this time the ○：Good ×：fracture

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 88

 becomes more advantageous. Fig.12 shows an Conclusions
example of the wrinkles. A blank holder force
greater than 1,400kN resulted in cracking in the A TRIP type, bainitic ferrite, 980MPa grade
stretch flange portion, as shown in Fig.13. The cold-rolled steel sheet with superior formability
developed steel, on the other hand, was formable was introduced. This sheet has improved elongation
without causing any cracking up to the equipment characteristics, compared with conventional type
capacity limit, a blank holder force of 2,000kN. Thus DP steel.
the developed steel can be used for increasing the The developed steel is superior in elongation
strength of parts and for the reduction of production characteristics and strain dispersibility compared
cost by single-piece forming. Furthermore, its with the conventional type DP steel and can be
excellent formability can be exploited to allow the applied to body frame parts (forming).
product design of parts having deeper cross-sections A formability evaluation, using a small laboratory
and to improve collision performance. scale die, has confirmed that the developed steel
has a formability exceeding that of conventional type
DP steel in all the terms of stretching, deep-drawing
and bending. It also exhibits a stretch-flangeability
almost equal to that of the conventional material.
Dies for large, actual-size parts simulating a B-pillar
and A-pillar were used to evaluate the formability of
the developed steel. The results confirm that the
developed steel has a formability that substantially
exceeds that of conventional type DP steel. The
developed steel can reduce costs by enabling the
single-piece forming of parts which otherwise had
been difficult to form and can increase the degree of
Fig.12 Example of wrinkle in test part freedom in improving collision performance.
(BHF＝1,300kN 980DP steel)

References
980 Developed steel 980DP steel
1) J. IWAYA et al., The Proceeding of The 36th Japanese Joint
Conference for the Technology of Plasticity, (1985), p.309.
2) J. IWAYA et al., R&D KOBE STEEL ENGINEERING
REPORTS, Vol.47, No.2 (1997), P.33.
3) M. NAKAYA et al., R&D KOBE STEEL ENGINEERING
Fracture
REPORTS, Vol.59, No.1 (2009), P.46.
4) M. HAKAYA et al., R&D KOBE STEEL ENGINEERING
REPORTS, Vol.57, No.2 (2007), P.19.
5) Japan Sheet Metal Forming Research Group, Press Forming
100mm 100mm Difficulty Handbook (in Japanese), Nikkan Kogyo Shimbun
Ltd., 2007, p.78.
Fig.13 Example of press result (BHF＝1,500kN) 6) O. MATSUMURA, Tetsu-to-Hagane, Vol.79, No.2 (1993), p.209

89 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

 Integrated Application Method for KOBEHONETSUTM
Steel Sheet
Dr. Yasuo HIRANO＊1, Tetsuya IGARASHI＊1, Haruyuki MATSUDA＊2, Makoto NISHIMURA＊2
＊1
Sheet Products Development Department, Research & Development Laboratory, Iron & Steel Business
＊2
Mechanical Engineering Research Laboratory, Technical Development Group

KOBEHONETSU is the trade name for a steel sheet its surface. When used for the cover of a home
having heat-radiation ability. A method was developed electronic appliance or of an office automation
for applying this steel sheet to dissipate heat in high- apparatus, for example, it efficiently dissipates the
performance electronic equipment. By applying internal heat to the outside by radiation heat transfer
KOBEHONETSU to both the heat sink and housing and lowers the internal temperature of the
simultaneously, the amount of heat transferred out of an instrument.
electronic device was significantly increased to a level The efficiency of radiation heat transfer is
comparable with that achieved by a cooling fan. This proportional to the fourth power of the absolute
cooling effect is enhanced by increasing the heat sink temperature. Therefore, the higher the internal
area, which makes KOBEHONETSU more applicable temperature of an instrument, the higher the heat
to heat sources such as CPUs. For example, this steel radiation efficiency. On the other hand, heat
sheet can be machined into heat sinks, while radiation works less efficiently when the ambient
maintaining its area and heat dissipation capability, temperature is low and the difference between the
for cases where heat sinks may interfere with other ambient and internal temperatures is small.
parts. This paper introduces a method of achieving a
heat radiation effect comparable to that of a cooling
Introduction fan, even when the ambient temperature is low and
the difference between the ambient and internal
An increasing number of home electronic and temperatures is small.
office automation products have become digitally
operated. In the digital circuits of these products, 1. Principle of heat radiation structure
most semiconductor devices work as on/off
switches. The on/off action of these semiconductor The amount of heat transferred by thermal
devices consumes electrical energy, but the greater radiation is considered using a semi-cylinder model
part of the energy supplied escapes as heat. Higher- consisting of a gray body having a heating element,
speed semiconductor devices consume more power, an enclosure and an outer space, which are located
increasing the amount of heat generated and raising in that order from inside out (Fig. 1) 1). In the figure,
the temperature. In other words, improving the the symbols ①, ②, ③ and ④ represent the heat
performance of a digital instrument is inevitably radiation surface of the heating element, the inner
associated with an increased amount of heat being surface of the enclosure, the outer surface of the
generated and a rise in temperature. enclosure and the surface facing the outer space,
A rise in the internal temperature of an instrument respectively. Their surface areas are represented by
can cause its semiconductor devices to malfunction, A1, A2, A3 and A4, while their emissivities are
cause the characteristics of its elements, such as represented by ε1, ε2, ε3 and ε4, respectively. The
resistors, to change, and shorten the overall life of surface temperature of the heating element is
electronic components containing organic insulation. denoted by T1, the inner surface temperature of the
To resolve these issues, heat dissipation technologies enclosure by T2, the outer surface temperature of the
have been developed for efficiently inducing the heat
generated inside an instrument to move away from
Environment
the heat source and exit the instrument. The heat
Q34
dissipation employs various techniques such as heat
sinks and fans.
As a manufacturer of steel sheets for the covers Q12
and chassis of home electronic appliances and
office automation equipment, Kobe Steel supplies
KOBEHONETSU, a steel sheet having heat ④ ③ ② ①
radiation ability 1). KOBEHONETSU is widely used Heat source Cabinet
as steel sheet with increased thermal emission from Fig. 1 Calculation model for thermal radiation transfer

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 90

 enclosure by T3, and the temperature at the wall by increasing the emissivity of the enclosure. The
surface facing the outer space by T4. Ignoring the heat following introduces the demonstration of these
transfer by thermal conduction and convection, the analysis results.
radiation amount Q12 (W) from the heating element
to the enclosure is given by equation (1). 2. Experimental verification of heat radiation
A1σ 4 4
（T1 −T2 ） structure
Q12＝ 1
             A1 1 …………………………(1)
＋ （  −1）
   ε1 A2 ε2 An apparatus, shown in Fig. 2, was made for
On the other hand, the amount of radiation Q34 (W) experimentally verifying the calculation results
from the enclosure to the outer space is given by obtained in the previous section. The enclosure,
（T34−T44）
A3σ having an opening at the front, was made of an
Q34＝
            
1 A3 1 …………………………(2) electrogalvanized steel sheet and had a dimension
＋ （  −1）
   ε3 A4 ε4 of 300mm(W)×88mm(D)×250mm(H). Vent holes,
where σ is the Stefan-Boltzmann constant having a each sized 100mm×30mm, were provided at the top
value of 5.667×10-8(W/m2K4). Assuming that T2 is and bottom to allow cooling by natural convection,
equal to T3 in equation (1) and equation (2), T1 is given the cooling method commonly employed by actual
as follows: instruments. A ceramic heater (1.8W), sized 25×25×
Q12 1 A1 1 2mm, mounted on an electronic board, was provided
T1＝ T44＋ ＋ （  −1）＋
A1σ ε1 A2 ε2 on the surface at the far end of the enclosure. This
1
                
Q34 1 A3 1 apparatus was used to compare the temperatures
＋ （  −1） 4 ………………(3)
   A3σ ε3 A4 ε4 of the ceramic heater in the following four cases:
equation (3) was used to study the effect on heating 1) The opening at the front was closed with an
element temperature caused by the surface area of electrogalvanized steel sheet (Fig. 3 (a))
the heating element and the emissivity of the 2) The opening at the front was closed with a
enclosure (Table 1). Here the following assumptions KOBEHONETSU sheet (Fig. 3 (b))
were made, i.e., the amount of heat generated is 0.3W 3) The opening at the front was closed with an
and the wall surface facing the outer space has an electrogalvanized steel sheet with a fan provided
area of 100m2, an emissivity of 0.99 and a temperature at the center of the sheet so as to cool the ceramic
of 25℃. Case 1 assumes that the enclosure is made of heater directly (Fig. 3 (c)) and
an electrogalvanized steel sheet. The heating element
temperature for this case is calculated to be 79℃. 88mm
300mm
Case 2 assumes that the enclosure is made of
KOBEHONETSU. The heating element temperature Upper vent hole

for this case is calculated to be 74℃. Case 3 assumes Electronic board

250mm

that the heating element has an emissivity equal to Heat source

that of KOBEHONETSU and has twice as much

surface area. This case also assumes that the 100mm
enclosure is made of KOBEHONETSU. In Case 3, the (a)
30mm
Lower vent hole (b)
temperature of the heating element is lowered to Fig. 2 Experimental apparatus for heat radiation structure
53℃. Case 4 assumes that the heating element has a
surface area three times larger. The heating element
temperature for this case is lowered to 44℃.
These results indicate that the temperature can be
lowered significantly by increasing the emissivity, as
well as the surface area, of the heating element and

Table 1 Influences of surface area and emissivity of heating (a) (b)

element and emissivity of housing on temperature of
heating element
A1 A2＝A3 Q12＝Q34 T1
Case 2 2 ε1 ε2＝ε3
(cm ) (cm ) (W) (℃)
1 10 0.8 0.1 79
2 10 0.8 0.86 74
1,200 0.3
3 20 0.86 0.86 53 (c) (d)

4 30 0.86 0.86 44 Fig. 3 Cooling method of heat source

91 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011

 100 Heat sink

90
Temperature (℃)

Radiation
80 Ambient temperature : 23℃

70
Radiation IC chip (3W in total)
60

50
Front panel : EG Front panel : KB Front panel : EG Front panel : KB Exterior cover
＋fan cooling ＋KB heat sink
Natural
Fig. 4 Effects of heat radiation structure convection
(EG : Electrogalvanized steel sheet, KB : kobehonetsu)
Board

4) The opening at the front was closed with a Fig. 5 Simulation model for heat radiation structure
KOBEHONETSU sheet, while the ceramic heater
was covered by a heat sink (40×40mm) made of 90
80
KOBEHONETSU (Fig. 3 (d))

Temperature of IC (℃)
70
As shown in Fig. 4, the results indicate that the
60
ceramic heater covered with a KOBEHONETSU 50
sheet exhibited a heat dissipation comparable to 40
that of the cooling fan. 30
20
3. Method for effectively utilizing heat radiation 10

structure 0
0 50 100 150 200 250 300 350
Area of heat sink (cm2)
As reported in the previous section, a ceramic Fig. 6 Effects of heat sink area on heat source temperature
heater, sized 25×25mm, covered by a piece of
KOBEHONETSU, 40×40mm, realizes a temperature
reduction of 17 ℃. The results shown in Table 1
also suggest that increasing the size of the Electronic part
KOBEHONETSU sheet placed on the heater should
lower the temperature even further. To verify this
Heat sink
and to establish a guideline for the size of a Heat source

KOBEHONETSU heat sink that will achieve the

desired temperature, an analysis was conducted
according to the simulation model shown in Fig. 5.
The results are shown in Fig. 6. The heat source
temperature decreases with the increasing area of a Electronic part

KOBEHONETSU heat sink. This suggests that the

heat sink area can be set up according to the target
Fig. 7 Constraints on heat sink due to other electronic
temperature.
components
There may be cases where the heat sink area
cannot be set up in a desired size, due to other
components located near the heat source (Fig. 7). In
such cases, the heat sink can be adapted to avoid
interference with neighboring components. Several
examples are shown in Fig. 8, and the results of the
corresponding temperature calculations are shown
in Table 2. The heat source temperature can be
decreased by increasing the heat sink area, while
avoiding interference with neighboring components,
as shown by Case (c) and Case (d). Even in a case
where a vertical surface exists in the heat source, as
in Case (d), a KOBEHONETSU heat sink is effective (a) (b) (c) (d)
in lowering the temperature. Fig. 8 Cases of setting up heat sink in case of parts close to
heat source

KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011 92

 Table 2 Temperature reduction by setting up heat sink in of electronic instruments without decreasing the air-
case of parts close to heat source tightness of the apparatus. KOBEHONETSU can be
Area of heat sink Temperature of heat source used not only for covers, but also for heat sinks, the
Case 2
(cm ) (℃)
combination of which decreases the temperature of
(a) None 103
the heating elements even further, allowing the use
(b) 9 79
of cooling fans with smaller capacities, or even
(c) 16 68
their elimination, which leads to cost reduction.
(d) 20 62 KOBEHONETSU is expected to further contribute
to the improvement of instrumental performance, in
Conclusions terms of performance upgrading, downsizing,
energy saving and noise reduction.
The interior temperatures of electronic instruments
have come close to the heat resistance limit of References
arithmetic elements such as CPUs. The life of the
precision components is in question. KOBEHONETSU 1) Y. HIRANO et al., R&D KOBE STEEL ENGINEERING
offers a method for lowering the internal temperature REPORTS, Vol.54, No.1 (2004), p.58.

93 KOBELCO TECHNOLOGY REVIEW NO. 30 DEC. 2011