Proceedings of The Twelfth (2002) International Offshore and Polar Engineering

Kitakyushu, Japan, May 26-31, 2002

Copyright 0 2002 by The International Society of Offshore and Polar Engineers
ISBN l-880653-58-3 (Set); ISSN 1098-6189 (Set)

Characteristics

Conference

of TMCP steels and their welded joints used for hull structures
Department

kfasahiro
Toyosada
of Marine Systems Engineering,
Fukuoka, Japan

Kyushu University

CLASSIFICATION

AND METALLURGY

OF TMCP STEEL

ABSTRACT
TMCP steels are basically classified into two types: non-accelerated
cooling (Non-AcC) and accelerated cooling (AcC) processes. Fig.1
shows a schematic diagram of TMCP in comparison with conventional
rolled process.

TMCP steels were developed in Japanese steel makers about 20 years

ago. They are extensively used in ship building industry and are
diversifying to other industries such as marine structures, bridges and
so on. Reentry European and Korean steel makers produce TMCP steel
also. They have not only great merits but also possible demerits. In this
paper, these merits and demerits are explained in addition to recent
topics in Japan about TMCP steel.

KEY WORDS: TMCP steel; low carbon equivalent; weldability;

toughness; fatigue strength; softening of HAZ; distortion; residual
stress

INTRODUCTION
For recent years, ships with extensive use of high tensile steels with
tensile strength of 5OOMPa class (HT-500) have increased more and
more in number in Japan for the purpose of reducing their hull weight,
thus leads to reduce its material cost. This is because ThermoMechanical Control Process (TMCP) has been developed and TMCT
steels have good weldability due to their low C,,
Though Japanese steel makers developed TMCP steels that meet
the extensive demand of strength and toughness for line pipe had
become a center of attraction from Japanese shipbuilders.
The
shipbuilders have utilized various highly efficient welding technologies
in their fabrication. Accordingly, a great deal of joint work among them
including university professors has been carried out to put such steel
and welding technology into practical use in the Shipbuilding Research
Association of Japan (SR193), that has resulted in the accumulation of
an abundance of useful data.
This paper shows the characteristics
of TMCP steels and their
welded joints from the viewpoints of welding procedures, fracture
toughness and fatigue strength, which is summarized mainly in the
SR193 committee (1985). Moreover recent topics about TMCP steels
are also explained.

Time
Fig. 1 Schematic illustration of thermo-mechanical
control
process (TMCP) (TMR: thermo-mechanical
rolling, AcC:
accelerated cooling, AC: air cooling)
Non-AcC process consists of (1) low slab reheating temperature
and (2) intensification
of rolling reduction in the austenite unrecrystallized region. The controlled rolling is finished either in the
region of austenite or in the intercritical region, austenite + ferrite
(r+a).
In AcC process, accelerated cooling is carried out after controlled
rolling. Cooling rate and finishing-cooling
temperature in the process
are controlled depending on required properties. Maximum available
thickness in AcC type TMCP steel is 100 to 200 mm at present.

385

less C,, than conventional steel.

Since TMCP steel is characterized by its low C,, (or low C) and its
fine microstructure,
some merits and demerits are considered to bring
up as follows.

MERITS OF TMCP STEELS

Since its low C,, and low C characterize TMCP steel, the hardnability and cold cracking susceptibility of HAZ of TMCP steel are
expected to be improved compared with conventional steels.
600
z
Ei
: 500
I
&

Fig.2 Microstructural changes of austenite and subsequent

transformation during TMCP

2
2 400
B
F
u
ii
1 300

Fig.2 shows the schematic metallurgical changes of TMCP steel

during the process from slab reheating to final cooling. The purposes of
lowering the slab reheating temperature are: (1) refining the initial
austenite grain size before rolling, (2) promotion of recrystallization
and expansion of the recrystallized region to a lower temperature, and
(3) grain refinement of recrystallized austenite grains by rolling in the
recrystallized region.
Intensification
of rolling reduction in the unrecrystallized
region
results in an increase in austenite grain boundary surface per unit
volume and density of deformation band in the matrix of austenite
grains. It greatly contributes to the refinement of the microstructure
after transformation.
In addition, accelerated cooling after controlled
rolling plays a role of clogging the growth of grain size.
Fig.3 shows an example of the relation between tensile strength and
C,,. It can be understood that TMCP brings about further improvement
in strength. The increment of tensile strength for TMCP steel comes
from the refinement of the microstructure, which can be predicted by
Petchs law. From the figure, we can make steels with low C,, by
TMCP compared with conventional process.
As for ship structures and offshore structures, 320, 360 and 400
MPa yield strength class HT-500 steels, which have experience to be
used for actual ship hull. Obviously, AcC type TMCP steel has quite

5
E
X
9 200
0

Fig.4

50
100
Bead length lmml

150

Relationship between maximum hardness numbers

and bead length

Fig.4 shows the results of maximum hardness test for HT-500 steels
with yield strength of 320MPa (YP320) and 360MPa (YP360) classes
in accordance with JIS maximum hardness testing. As seen in the figure,
the maximum hardness decreases as C,, is lowered and bead length
becomes large. Since it has the possibility of generation of microcracking in the HAZ if the maximum hardness becomes larger than
400Hv or more, JSQS (Japan Ship Quality Standard) exhibits the short
bead with length less than 50mm for conventional HT-500 steel. From
the figure, short bead with length greater than 1Omm can be applied for
YP320 and YP360 AcC type TMCP steels of which C,, is generally
less than 0.36%

6oo-

\z?-TMCP YP32d . YP360 steel Plates

0.3

0.25

0.35

C,, = (C+Mn/6+Si/24)

Fig.3 Relationship

Bead length [mm]

0.4
[%]

Fig. 5 Relationship

between C,, and strength

386

between H,,,

and welded bead length

Fig.5 (Machida et al, 1988) is the relation between the maximum

hardness in the HAZ and bead length for YP320 and YP360 AcC type
TMCP steels. The YP400 class steels are nearly equal with the NonAcC type TMCP manufactured YP320 and YP360 steels as regards the
hardening of HAZ. This may be because the C,, upper limit remained
almost same.
From the viewpoint of maximum hardness being not in excess of
400Hv, bead length may be allowed up to 30mm for YP400 AcC type
TMCP steel, and YP320 and YP360 Non-AcC type TMCP steel.
I
I
,100
P
0
t

be said that in actual welding operation using low-hydrogen electrodes,

preheating may be unnecessary at the ambient temperature of over 0C.
Moreover the application of medium hydrogen electrodes becomes
possible as shown in Table 1.
Table 1 Weld cracking susceptibility
-

NO

stee

rhdaess

[mm1

of TMCP steels

Hydrogen
content
[rn/lOOg]

-1

3M-1

20

0%
00

0.26

0.3

0.34

3M-2

0.38

C,, (WES)=C+Si/24+Mn/6+Nil40+Cr/5+Mol4+V/14

iH36

[%]

1M-3

IH36

14

20

01
0

'
10

20

'

30

Heat input

'

40

'

50

,M-4

H36(

35

20

[MJ/m]
-

Fig.7 Relationship between heat input and HAZ toughness at fusion

line in comparison with AcC and non-AcC type of TMCP steel
plates and conventional steel plates

N no crack

(**) (I) At startmg end

(II) At mddle
(III) At timshmg

While many methods of cold cracking test have been proposed, JIS
small Y-groove cracking test is used here for summarizing the data on
the cold cracking susceptibility of the HAZ for HT-500 steels. Fig.6
shows the effect of C,, on the critical preheating temperature to prevent
cold cracking of conventional
steels and AcC and Non-AcC type
TMCP steels by using small Y-groove cracking test.
As the small Y-groove cracking test specimen has very large
constraint, the critical preheating temperature to prevent cold cracking
in the small Y-groove cracking test is higher by about 75C than that
applicable to welded joint in general steel structures. Therefore it can

end

The reduction
of the C,, served effectively
for toughness
improvement of the HAZ, particularly when high heat input welding is
used. This is generally proved by the fact that the welded joint
toughness of mild steels is superior to that of high tensile steels. This is
due to the formation of fine ferrite by the lowered hardenability as well
as reduced volume fraction of island martensite and also reduced
amount of cementite. However, the reduction in C,, generally brings
about a reduction in strength of the plates. Therefore, in conventional
manufacturing methods such as as-rolled or as-normalized processing,
efforts were made to adjust the MnK ratio or to replace C by Ni or Cu

387

to suppress C,, as low as possible. By these techniques, however, no

substantial reduction of C,, would be achieved to improve HAZ
toughness. As mentioned previously, the reduction of C,, can be
achieved by applying TMCP.
Fig.7 shows an example of the relationship between heat input and
HAZ toughness at fusion line in comparison with AcC and Non-AcC
type TMCP steel plates and conventional steel plates. AcC type TMCP,
in particularly, brings about further improvement in toughness and
makes possible to be applied for very high heat input welding. It is
natural that the improvement of HAZ toughness leads to increase
fracture toughness such as critical CTOD and Kit for the welding joints.

HAZ

::<I

fusion
line

Base

grains and intercritical region is the second deteriorated part. On the

other hand, in the TMCP steel treated by adjusting microalloying and
gaseous elements, typical deterioration is not observed after single
thermal cycle. However after 2nd thermal cycle, new deteriorated
microstructure generally generates in the intercritical region for 2nd
thermal cycle at the coarse grain HAZ due to the first thermal cycle.
This sudden deterioration
is caused by the formation of the
substructure
named
martensite-austenite
constituent
(M-A)
or
martensite islands. This deteriorated area is very small and is scattered
around in the HAZ near the fusion line. Succeed thermal cycle of which
peak temperature is over about 400 degree decomposes martensite
islands, and then the toughness
recovers.
Although
there are
deterioration parts in toughness in the HAZ of TMCP steel, the HAZ
toughness of TMCP steel is further high compared with that of
conventional steel as shown earlier in Fig.7.

metal

HAZ
bondary
a) Afler single thermal cycle

HAZ

:j:

- Base metal

g------
f,,,
.Itli
./I- L
I

TMCP steel i

I
I
I
I
I
I
I
j

I coarse :
:
I

grain
realm

fusion
line

I
I

fine grain
rwon
;

Temperature

1 / TK [x 1 Om3K]

I
I
I
I
I
I
I
I

I
1A,,

-Lnt

Fig.9 Comparison

of crack arrestability

Moreover it is observed that Non-AcC type TMCP steel plates,

especially the plate manufactured by TMCP which the controlled
rolling is finished in the region in the intercritical region ( y + a ), have

ion.1 steel

excellent brittle fracture arrestability

as shown in Fig.9.

. intercritical region

through the refinement

I
- IP

HAZ
bondary
a) Afler single thermal cycle

Fig.8 Schematic illustrations about changes of fracture toughness in

HAZ near the fusion line

By the way, it is well known that many different microstructures

generate in HAZ depending on thermal cycles due to welding and
chemical composition
of steel. Fig.8 shows the typical schematic
changes of toughness in HAZ near the fusion line considered from the
Charpy test results for the plate with simulated welding thermal cycle.
Fig.Sa) shows the schematic changes of toughness after single thermal
cycle. Generally speaking, HAZ near the fusion line is the most
deteriorated part in conventional steel due to the generation of coarse

Microstructure

Mmostructure

in Surface Layer

with Ultra-Fine

grains (SIJF)

Photo 1 Microstructure

388

in Mid-thickness

of SUF steel

of grains

The softened HAZ may affect the buckling

tensile strength. However if the softened HAZ
highly strained zone such as plastic hinge lines
transverse direction in the middle of the panel
the bucklina strenath due to the softened zone is

By the way, temperature at the layer close to plate surfaces rises

after stopping AcC with high intensity of cooling. Grains of
microstructure after rolling in the intercritical region under temperature
rising process become finer than under temperature cooling process.
Then Rolling technique after finishing AcC had been studied and plates
with layer with quite fine grains of 1 to 3 b m close to the surface, as
shown in photo 1, are then developed. They are called SUP steel plate
and show ultra high arrestability for brittle crack propagation (Ishikawa
et al, 1997)

POSSIBLE

DEMERITS

strength as well as the

is located away from
that are located in the
plate, the reduction of
nealiaiblv small.

OF TMCP STEELS

Since TMCP steels obtain sufficient strength and toughness without

heat treatment, the problems in strength will mainly create due to
reheating such as by hot working and PWHT. Although
some
countermeasures for the problems due to hot working and PWHT have
been already prepared, the explanation about these is intended to delete
here because hot working and PWHT are usually not applied for ship
hull structures.
TMCP steel plates, due to their low C,,, tend to decrease their
welded joint strength through the softening of HAZ caused by high heat
input welding with heat input greater than about 70kJ/cm. On the
contrary the softening of the HAZ does not appear in the normal steel
because of reinforcement by alloying elements.
However, the decreased HAZ strength stays at a level of about 90%
or more of that of the base plate and the width of the softened HAZ is
less than about 70% of the plate thickness, even when the very high
heat input welding, say 300kJ/cm, is applied. This comes from the
restriction of lower limit of C,, as shown in Table 2.

CO,

0.1
IO4

sm-automatic

CO,

Gas

shlelded

IO5
IO6
Number of cycles to fracture

Arc Process

IO'
: N,

Fig. 11 S-Nf curves (Base metal, butt welded joints)

Table 2 Range of C,, for HT-500 steel

Therefore, no problems will arise in securing the required strength.

Moreover if the width of the tensile test specimen is 10 times the
specimen thickness, the joint strength is far greater than that of the
standard small width specimen as shown in Fig.10.

5~0

AcC type
(AH320.

iMCP
C,,=O

steel
25%. t=25mm)
Solid marks

Flush

520

;5001+-----? 1
iYl
a,
=

NK rule ( 490 [MPa] for YP360

Stress intensity factor range : A K [MPa m]

Fig. 12 Relationship between crack growth rate and stress intensity

factor range

steel )

2 480
P

a
________________________________________-----The existence of the softened HAZ in case of high heat input
0
NK rule ( 490 [MPa] for YP320 steel )
welding is also considered to decrease its fatigue strength. Fig. 11 shows
460 ________________________________________---------------------------an example of fatigue test results obtained from YP400 AcC type
: Tensile strength of softened HA2
TMCP steel and its butt welded joints by CO2 gas shielded arc welding
I.
I-.
I.
I.
I
I
200
300
400
0
100
and FCB welding. The heat input for the FCB welding was 149kJ/cm.

The vertical axis in the figure represents

obtained by dividing the cyclic nominal
metal tensile strength oB. The solid line
the mean fatigue strength and the range

Specimen width , W [mm]

Fig. 10 The effect of plate width on the tensile strength of

welded joint

389

a dimensionless stress range

stress range oR by the base
and the dot dash lines show
of data scatter for CO2 gas

welded joints and FCB welded joints of conventional HT-500 steels

respectively.
Similar results were obtained for YP320 and YP360
TMCP steel welded joints. It is obvious that the effect of softened HAZ
on fatigue strength for TMCP steel is almost the same as that for
conventional steel with the same tensile strength.
Fig. 12 shows the test results of fatigue crack propagation rate in the
softened HAZ for AcC type TMCP steel. These joints were prepared by
very high heat input welding as shown in the column in the figure. In
this figure, the data for conventional steel welded joint is also included.
From the figure, it can be also said that the fatigue crack propagation
rate in the softened HAZ is almost the same as that in the HAZ for
conventional steel.
As explained above, the effect of the softened HAZ on fatigue
strength of the joint is negligibly small. However it is well known that
the fatigue limit of high tensile steel does not increase depending on its
static strength at highly stress concentration area. So we must take care
of this factor when high tensile steel is used. From the point of view,
when high tensile steel, especially YP400 steel, is used, designers must
carry out the detail stress analysis such as 3-D FEM and higher grade of
detail design standards must be applied to reduce stress concentration
of local areas where considered critical, such as (1) bracket end of
bottom transverses, (2) openings, holes, cut-outs, (3) radius comers at
connections, (4) toes of tripping brackets, and (5) tapering of faces in
transition arrears.
In the manufacturing process of AcC type TMCP steel plate, if the
uniformity of temperature within the plate can not be maintained, flame
cutting a plate into strips releases residual stress, thus leading to
distortion as shown in Fig. 13.

under development
with expectation
of realization
construction with high accuracy-of its shape and scale.

for

block

loo80
F
60
.-E
P
g

40

20

.k

0I

Steel 1 Steel 2 Steel 3 Steel 4 Steel 5

I
I
Usual TMCP

Fig. 14 Comparison

I
Developed

I
TMCP

of measuring deflection after flame line heating

100
z
.2
g 50
m
5
6
0
8

0.4

7
&.P E
540.2
$E
mE
(I)0
z 200

Fig. 13 An example of deflection after plate stripping

B
5
5
-0 100
2
b
2
Q
0

To prevent this, high finishing cooling temperature is usually

applied in addition to uniform cooling by controlled water pouring
distribution for width direction of steel plate. This technique has been
already established. But engineers in assembly sections on shipbuilding
factories have been feeling empirically and vaguely that TMCP steels
have larger scatter of distortion after thermal processes such as flame
cutting and welding than conventional steel. Cold leveler with high
press are developed for removing residual stress in TMCP plates after
compressive yielding in almost half zone of a plate in thickness
direction (Tani et al, 2001). Fine control of alloying elements, rolling
timing and cooling rate in TMCP is of cause achieved for the purpose
of getting sufficient strength, toughness and weldability.
Fig.14 (Tani et al, personal communication)
is an example of
testing results of deflection after flame gas heating on lines parallel to
the diagonal of 16mm thick steel plate with 2500mm width and
3000mm long. Heating pitches of 300mm were applied for both of
normal AcC type TMCP steels and above developed TMCP steels.
Heating conditions are the same. It is seen that scatter of deflection is
very small each other in the developed TMCP steel after the reinforced
cold leveler in comparison with usual TMCP steel. This steel is now

-100

0
Tempetture

100

2 IO

[Deg]

Fig. 15 Results of V-notch Charpy impact test

The last possible demerit is concerning about separation. Because
of the enhanced strength of TMCP steel by controlled rolling, so-called
separation or splitting is often observed in the fractured specimens
taken from the longitudinal or the transverse direction of steel plates,
especially in steel plate controlled rolled in the intercritical temperature
region. Cracks called separation propagate in planes perpendicular to
the main fractured surface and parallel to the plate surface during the
final fracture process due to restriction force exposed in the thickness
direction. Separation starts from an origin such as micro-orientation
texture boundary or an elongated non-metallic inclusion such as MnS,

390

which would be formed or elongated during TMCP rolling. Separation

is particularly observed in the fracture surface of V-notch Charpy
impact specimens.
Separation index is usually used for the purpose of quantifying the
degree of separation, which is defined by
SI =Cei/A

(end>

Fig.16 shows the comparison

between
the plate thickness
directional toughness and longitudinal directional toughness in CTOD
tests. The horizontal axis and vertical axis in the figure show a
temperature where the critical CTOD becomes equal to 0.2mm for the
longitudinal and thickness direction in CTOD tests respectively. The
small triangle marks are for AcC type TMCP steels of which V notch
Charpy specimens showed no separation. The small circular marks are
for conventional
HT-500 steels. Ductility in the plate thickness
direction has a mutual relation with sulfur content. So in the figure, the
range of sulfur content of steels used is also shown.

(1)

where SI: separation index (mm/mn?),

!, : Each separation length (mm),
A: Area of main fracture surface.
Fig. 15 shows an example of Charpy transition and separation index
curves for a Non AcC type TMCP steel. Separations are observed only
in the transverse or the longitudinal directional specimens. Separation
depends clearly on the difference of toughness transition phenomena
between transverse (or longitudinal) and plate thickness directions. The
separation index indicates a maximum value at a temperature where the
difference in toughness between in transverse (or longitudinal) and in
plate thickness direction becomes evident as shown in Fig. 15.

It is obvious from the figure that maximum separation index does

not depend on sulfur content. It can be also seen that the degree of the
deterioration in critical CTOD in plate thickness direction compared
with in longitudinal direction for TMCP steels is almost the same as
that for conventional steels except for the steel having excessively high
maximum separation index of 0.6-0.7mm/mm2.
Then it comes to
conclusion that clean steel with a maximum separation index of
0.5mm/mm2 or less may not create any problem in its practical
application.

CONCLUDING
5 O-

TMCP steels are excellent in terms of steel properties and ease of

fabrication as compared with conventional steels. However there are
some possible demerits. Then we must use these steels with carefully
consideration about these demerits.

OH
0
F
.g
8

REMARKS

REFERENCES
-5 O-

Ishikawa, T., Imai, S., Inoue, T., Watanabe, K., Tada, M. and
Hashimoto, K., (1997).Practical Assessment of Structural Integrity
of Ships Attained by the Use of SUF Steel Having CrackArrestability
(SUF: Surface-layer with Ultra-Fine grain) , 16th
OMAE, Vol.111, p.301-308,
Machida, S., Kitada, H., Yajima, H. and Kawamura, A. (1988).
Extensive Application of TMCP Steel Plates to Ship Hulls 40
kgf/mm Class Yield Stress Steel, Int. J. ofMarine Structures, Vol. 1,
No.3
SR193 committee
(1983,1984 and 1985). A study of effective
application for TMCP steels, Shipbuilding Research Association of
Japan, No.367 (1983), No.374 (1984) and ISSN 050-1480(1985)
Tani, T., Okada, N., Ohe, K. and Miyazaki, M. (2001). Effects of
Residual Stress Controlled TMCP Steel Plate on Accuracy Ship
Blocks, J. Sot. Naval Architects of Japan, Vo1.189, p.299-307

6
I
t?
t -10 Og

-15 OL
-100

-50

50

TsCC02(
L - Direction ) [Deg]

391