Mechanical Properties and Retained Austenite

in Intercritically Heat-Treated Bainite-Transformed

Steel and Their Variation with Si and Mn Additions
YASUHARU SAKUMA, OSAMU MATSUMURA, and HIROSHI TAKECHI

Processing peculiarities and functions of alloying elements, such as Si and Mn, were studied
for improving formability of steel sheets with mixed microstructures. Annealing a sheet steel
with 0.2 pct C in the intercritical range produced very fine particles of retained austenite which
were moderately stabilized due to C enrichment by subsequent holding in the bainite transfor-
mation range. Its strength-ductility balance is greatly superior to that of other dual-phase steels
due to transformation-induced plasticity (TRIP). The holding time in the bainite transformation
range varies with temperature, depending on the activation energy of C diffusion in austenite,
and shifts to longer times with an increase of Si or Mn additions. The optimum cooling rate
from the intercritical region is reduced with an increase of Mn content but is not influenced by
Si content. Additional Mn makes the retained austenite content larger, although uniform elon-
gation remains the same. In this case, the product of tensile strength and total elongation is
increased due to an increase in the tensile strength. Contrary to Mn, Si does not affect retained
austenite content but improves the uniform elongation by increasing its stability.

I. INTRODUCTION tensile strength of about 980 MPa and a total elongation

larger than 30 pct can be obtained. Attempts to under-
IT is well known that dual-phase steels til have high stand the role of Si in this process t23j have already been
strength and ductility combinations for steel sheets ap- tried, but the details involving the influence of Mn t241are
plied to press forming. Further improvements in strength still ambiguous. Also, the feature of intercritical an-
and ductility can be achieved by transformation- nealing and the effects of the chemistry and heat-treatment
induced plasticity (TRIP) of retained austen- condition on the resulting mechanical properties have not
ite. t2.3,4] Transformation-induced plasticity was the been investigated sufficiently. This paper intends to clear
phenomenon first found in steel with large amounts of up the noticeable phenomena mentioned above using
Ni and Cr by Zackay et al.t51 However, large additions
0.2 pct C steel with different levels of Si and Mn contents.
of costly alloying elements and the accompanying
troublesome process make it difficult to employ this phe-
nomenon in manufacturing cost-effective steel sheets for
automotive use. It is also well known that a significant II. EXPERIMENTAL PROCEDURE
amount of retained austenite can be obtained in steel One-hundred-kilogram vacuum-melted ingots with a
containing about 2 pct Si by cooling to room tempera- chemistry and estimated transformation temperatures Act
ture through the two-stage bainite transformation re- and Ac3,1251 given in Table I, were prepared. Slabbed
gime. [6-91 Many studies 1~~ have been conducted in order materials 30-mm thick were reheated to 1200 ~ for 1 hour
to obtain a better understanding of the essentials of bain- and rolled in five passes to 4.5 mm by finishing at 900 ~
ite transformation, and recently, the latter has been used The hot-rolled sheets were air-cooled to room tempera-
to try to improve the toughness of steelf115-19] Reports on ture, so the microstructure was composed of ferrite and
the application of this phenomenon to improve the form- pearlite. After surface grinding to 3.5 mm, the sheets
ability of sheet steels were rare; t2~ however, the au- were cold-rolled with a reduction of 77 pct. These cold-
thors t21'221 found that in steel sheets with comparatively rolled sheets, with a thickness of 0.8 mm, were heat
low carbon and silicon contents, larger amounts of aus- treated in the cycles illustrated in Figure 1. Micro-
tenite could be retained, and formability was improved structure and mechanical properties were measured.
when rapid cooling was performed not from a single aus- For most of the experiments, heating was done by two
tenite range but from the intercritical range. The treat- salt baths (in Sections III-A through Ill-C), but a direct
ment, which is compatible with existing continuous heating heat-cycle simulator was used in the experiments
annealing lines, produces about 20 pct retained austenite demonstrating the influences of cooling rates (in
susceptible to TRIP, and cold-rolled steel sheets with a Section Ill-D). The heat cycle is constructed by an-
nealing in the intercritical range for 180 seconds, rapidly
cooling directly from the annealing temperature to the
YASUHARU SAKUMA, Visiting Scientist, Department of bainite transformation temperature range, and then hold-
Metallurgical and Materials Engineering, Colorado School of Mines, ing at the terminal temperature for a given time between
Golden, CO, is with the Sheet and Coil Laboratory, R&D-II, Nippon 0 and 7200 seconds before cooling to room temperature.
Steel Corporation. OSAMU MATSUMURA and HIROSHI TAKECI-H
are with the Sheet and Coil Laboratory, R&D-II, Nippon Steel
The annealing temperature in the intercritical range was
Corporation, Sagamihara 229, Japan. set to give equal fractions of austenite and ferrite. The
Manuscript submitted January 3, 1989. definite numbers according to the chemical composition

METALLURGICALTRANSACTIONSA VOLUME 22A, FEBRUARY 1991--489

 Table I. Chemical Composition (Weight Percent), Estimated Transformation
Temperature, and Intercritically Annealed Temperature (*C) of Steel Used
Steel C Si Mn P S Sol. AI Total N Act Ac3 TI
A 0.201 1.240 0.950 0.0040 0.0063 0.040 0.0023 749 849 790
B 0.200 1.560 0.951 0.0030 0.0057 0.041 0.0029 758 863 800
C 0.199 2.100 0.940 0.0030 0.0058 0.041 0.0027 774 887 810
D 0.203 1.240 1.250 0.0050 0.0062 0.042 0.0032 746 840 780
E 0.195 1.210 1.550 0.0050 0.0058 0.040 0.0020 741 831 780
F 0.202 0.810 1.870 0.0080 0.0065 0.038 0.0028 726 804 760
G 0.201 1.240 1.800 0.0030 0.0062 0.041 0.0028 740 823 770
H 0.202 1.640 1.834 0.0030 0.0062 0.038 0.0024 751 839 780

are given in Table I. The heating rate was nearly 100 ~ III. EXPERIMENTAL RESULTS
The cooling rate was about 150 ~ in the treatments
with salt baths and was varied in the range of 10 to A. Peculiarity of Intercritical Annealing Compared
130 ~ with a simulator. The holding temperature at with That in the Single Austenite Range
the bainite transformation range was set between 350 ~ The difference between intercritical annealing and an-
and 450 ~ Besides, some specimens were annealed in nealing in the single austenite range was investigated for
the single austenite range, followed by rapid cooling and steel E, designated in Table I. Variations in the me-
holding for a short time at 300 ~ or 400 ~ in order to chanical properties and the retained austenite content with
investigate their difference from typical structures of lower a holding time of 300 ~ or 400 ~ are summarized in
or upper bainite. Figure 2. When the steel is rapidly cooled to the bainite
Microstructures were identified by optical and trans- transformation temperature from the single austenite range
mission electron microscopy and X-ray diffraction mea- 860 ~ its retained austenite content is under 6 pct and
surements. Optical microscopy measurements were made decreases with holding time. The mechanical properties
on the specimens prepared by the tint-etching method, are mostly determined by the holding temperature and
with a solution of 4 pct HNO3 + 7 pct (NO2)3CtH2OH + are not strongly influenced by holding time. Tensile
saturated Na2S203, [261 and the point-counting technique
was employed to estimate the fractions of ferrite, bain-
ite, and retained austenite plus martensite. The X-ray
diffraction method was used to quantify the amount of 1500
retained austenite, and the calculations were based on ~"""W~V r7 ~7
1300 860~ 300~
the integrated intensities of 200t~, 211a, 220% and 311"}'
diffractions of Mo K~ radiation on a chemically polished
layer at a quarter thickness below the surface. This mea- 1100 A ~ A ~ 8 6 ^ 0~ ~ A
surement was performed also for samples with different co
F- 900
amounts of deformation. Tensile properties were deter- O~Z)~,~ 780~ 400 ~
mined with JIS-5-type specimens, 25-mm wide with 700 r y't-o-r-q--o
50-mm gage length. The measurement was conducted at
room temperature with a crosshead speed of 0.17 m m / s
in a hard machine.
,-~ 30
..... 9 ....... ~ ....... 0 ~ 0
"OE 20
I ntercritical annealing
E (T1 ~ 180s) u/w A ..... =.-,,::::::::::::::::::
~ 0 I I I [ I I I ,,I

Holding at bainite 20
transformation range o~ 15
(350"--450 ~ 0 ~-7200 s) ffl

~ 10

1oo~ 0:o 0 I
2O 50 100 200 500 10002000 500010 )00
Holding time (see)
Temperature Fig. 2 - - C h a n g e s in mechanical properties and retained austenite content
with holding time at bainite transformation range after annealing at
Fig. 1--Heat-treatment scheme. single austenite or intercritical range (steel E).

490--VOLUME 22A, FEBRUARY 1991 METALLURGICAL TRANSACTIONS A

strength (TS) ranges from 1000 to 1500 MPa, and total the bainitic region. Adjacent to the retained austenite,
elongation (T.E1) is about 10 pct; the product is com- martensite is found in specimens of shorter holding times,
paratively low, indicating a common balance between such as 30 seconds (Figure 4(a)), but almost disappears
strength and ductility. when the holding time is long enough to give the max-
However, mechanical properties show dramatic changes imum value in U.E1. As the time increases to as long as
with holding time in samples intercritically annealed and 2400 seconds, the amount of retained austenite is greatly
isothermally held in the vicinity of 400 ~ due to vari- reduced and coarse carbides are clearly recognized in the
ations in microstructure. The largest amount of retained bainite structure (Figure 4(c)). However, these carbides
austenite at room temperature is about 15 pet, although are hardly detected in specimens held for 240 seconds at
this does not correspond to the highest elongation. Ten- 400 ~ (Figures 3(c) or 4(b)).
sile strength decreases with holding times up to
240 seconds but does not change as much beyond this B. Effect of Isothermal Holding Time on Mechanical
time. The change of T.EI reflects that of uniform elon- Properties and Its Variation with Si and Mn Contents
gation (U.E1), and holding for about 240 seconds brings
the largest value of T.E1 to about 34 pct. In this in- Annealed in the intercritical range and cooled rapidly,
stance, TS is about 760 MPa; hence, the product of TS all kinds of steels examined here show changes of U.E1
and T.E1 exceeds 26,000 MPa.pct. and retained austenite content as a function of isothermal
The microstructures after typical heat cycles can be holding time similar to those presented in Figure 2.
compared in Figure 3 which shows electron micrographs Figure 5 gives other examples of holding at 400 ~ for
of the specimens held at a balnite transformation tem- steel A (the poorest Si and Mn contents), C (the richest
perature for 240 seconds. In the case of annealing in the Si content), and G (the richest Mn content). The changes
single austenite range, the so-called upper bainite of true stress, work-hardening rate, and retained austen-
(Figure 3(a)) or lower bainite (Figure 3(b)) constitutes ite content during tensile deformation were examined in
the largest fraction. Compared to this, a complicated a specimen which attained the maximum U.E1. The re-
microstructure mixed with ferrite, bainite, and retained suits for steels C and G are shown in Figure 6 which
austenite is formed by rapid cooling following inter- indicates the decrease of retained austenite with the
critical annealing (Figure 3(c)). Retained austenite in this progress of deformation, typical of the TRIP phenomenon.
structure is blocky and no larger than about 1 ~m, and Figure 7 presents the time required to obtain a max-
it is mostly located at grain boundaries. The bainite re- imum in U.E1 and retained austenite content. These data
gion does not contain carbide precipitates and so resem- show that with increasing additions of Si or Mn, longer
bles commonly termed bainitic ferrite. times are needed to obtain the largest amount of retained
Figure 4 summarizes the microstructural changes of austenite or U.E1. An important observation is that the
intercritically annealed specimens with holding times in time required to reach the maximum elongation is much

(a) 860~ 400~ (b) 860~ --* 300~ (c) 780~ --~ 400~
[ I

1/~m
Fig. 3 - - ( a ) through (c) Typical nficrostructure obtained by holding at bainite transformation range for 400 s after annealing at single austenite
region or intercriticaI range (steel E). In (c), the arrow points to retained austenite.

METALLURGICAL TRANSACTIONS A VOLUME 22A, FEBRUARY 1991--491

 (a) Martensite (30 s) (b) Retained austenite (240s) (c) Bainite (2400s)
I I
t/tin
Fig. 4 - - ( a ) through (c) Changes in microstructure with holding time at 400 ~ followed by intercritical annealing (steel E). In (a), the arrow
points to martensite, and in (b), it points to retained austenite.

Fig. 5 - - C h a n g e s in U.EI and retained austenite content with holding Fig. 6--Changes of true stress, work-hardening ratio, and retained
time at 400 ~ and the variation with the difference of Si and Mn austenite content with progress of deformation and the variation in-
additions. fluenced by alloying elements.

492-- VOLUME 22A, FEBRUARY 1991 METALLURGICAL TRANSACTIONS A

 holding times for steels A through C at 350 ~ in spite
2000 1.8 M n 0.9 Mn 1.2 Si of the wide differences in their Si contents. Also, in
v
171 ,# steels D and E, U.EI reaches its maximum value earlier
::5 1000
U. E / I - I /
[] than in steel A with a leaner Mn content. The figure shows
~) 500 that a linear correlation exists between the time it takes
E= U. E I J for U.EI to reach its largest value and the reciprocal of
E 200 [3/[]
E

.e>
100

50
#E~ 0"
S XE' / absolute temperature of isothermal holding, though there
is a discrepancy in the case of higher alloying of Si or
Mn or at a lower holding temperature.

o C. Influence of the HoMing Temperature on the

== 20
I I I I I I Resulting Properties and Its Variation with Si and
I-- 0.5 1,0 1.5 2,01.0 1.5 2.0 2.5 0.5 1.0 1.5 2.0 Mn Contents
Si content (wt.%) Si content (wt.%) Mn content (wt.%)
Tensile strength, U.E1, T.E1, the product of TS and
Fig. ?--Changes in holding time at 400 ~ to attain maximum con- T.E1, and the retained austenite content obtained by
tent of retained austenite or U.E1 with the difference in Si and Mn
additions.
holding each steel for the time needed to produce the
largest U.EI at each holding temperature were calculated
from the experimental results, similar to Section B, and
greater than that needed to reach the maximum amount are shown in Figure 9. The retained austenite content
of retained austenite. This lag is also widened if the ad- obtained by holding steel at 350 ~ is between 8 and
dition of Si or Mn is increased, with the former showing 12 pct, regardless of its Mn content. In the case of small
a more remarkable effect. additions of Mn as 0.9 pct (steels A through C), the re-
Changes in holding time to give the largest U.E1 are tained austenite content remains almost constant with in-
plotted in Figure 8 vs varied holding temperatures. Above creasing temperature up to about 425 ~ with higher
375 ~ longer times are needed to obtain maximum U.EI temperatures causing a sudden decrease (Figure 9(a)). If
with increasing additions of Si at constant Mn or vice the Mn content is increased to 1.2 pct, the amount of
versa. At 350 ~ the effect of time for the different retained austenite goes through a maximum with increas-
amounts of Si or Mn addition becomes complicated. For ing holding temperatures (steel D in Figure 9(b)). Fur-
example, maximum U.E1 can be obtained at the same ther addition of Mn makes retained austenite increase
continuously with holding temperatures up to 450 ~
(steels E and G in Figure 9(b)). Figure 9(b) also shows
that the peak content of retained austenite increases with
1.4 450 higher Mn additions. On the contrary, the amounts of Si
b ~ O ~ ' ~ % ~ i " : Mi n do not affect the retained austenite content at the max-
imum U.E1 (Figure 9(a)). This tendency holds true ir-
400
1.5 respective of holding temperature and different Mn
contents.
F(0.8Si)~ (])": It is widely accepted that the strength of bainite de-
1.6 350 creases as the transformation temperature is raised. Re-
I I I I ,I I flecting this rule, in steel with small amounts of Mn, TS
D decreases with a rise in the holding temperature (steels A
qT'-.......,, ~IV%. 0.9 Mn - 450 V
through D in Figure 9). However, in steel with large
~, 1.4 - %~ ~b
L
-1 amounts of Mn, like steel G in Figure 9(b), TS increases
B ( 1 . 6 S ~ ) ~ C ( 2 . 0 S i) (o when the holding temperature is raised above approxi-
\ ,,....,. 400 mately 400 ~ This tendency is due to the much in-
x~ 1.5 -

~z~xl y E creased retained austenite that replaces bainite; the strength

increment caused by strain-induced transformed mar-

t
A(1 2.Si) ~ x~l~"
350 tensite compensates for the strength decrement of bainite
I I I ..... I I I I "o
-5 formed at higher transformation temperatures.
3= Uniform elongation and T.E1 show the same pattern
1.2 S i 450
1.4 -v""'"~-"'~'~'~UX~E(1.5 Mn) of change: they increase with a rise in holding temper-
ature but decrease after reaching the maximum near 400 ~
\ \\ ~,_G(1 8an) 400
to 425 ~ The maximum value of U.EI achieved in-
creases with increasing Si at 0.9Mn (Figure 9(a)). In ad-
dition, the temperature that realizes maximum U.E1 shifts
from 400 ~ for steel A, which has the poorest Si and
16 # 350 Mn contents, to 425 ~ for steels B and C, which con-
I I I , I I I I tain greater amounts of Si. When steel is alloyed with
20 50 100 200 500 1000 2000 more Mn, the temperature is also raised to 425 ~ but
Time to obtain the maximum U.EI (sec) the obtainable U.EI is saturated at 1.5Mn (steel E)
Fig. 8--Changes in holding time to obtain the maximum U.E1 with (Figure 9(b)). Higher levels of Mn reduce U.E1; steel G
isothermal holding temperature. is lower than E, as recognized in Figure 9(b). At 1.8Mn,

METALLURGICALTRANSACTIONSA VOLUME 22A, FEBRUARY 1991--493

 1100 40 30000 -~ 20

UJ

1000 27500 _C(2.0 S i)

i ,~
#_ I iy-.u.~
30f ~12
I / .n:
9 " t"

800
C :5 .El

700 20000 A(1.2 Si)'Sa

A 20i" A ~

600 I I ~ I I i I r J i 17500 I I i J I rr 0~ I K I I I
350 400 450 15 350 400 450 350 400 450 350 400 450
(a) 0.9 Mn Holding temperature (~

100r [,0 30000

t 20f
I G(1.8Mn) w I G/Q
35 T.EI IE(15M} )<~\ E | ~ )2E

P
(1.2Mn)~ IJ~" 13

",1 i 4L\
350 I 400 450 1 5 ~ 1750013~
0 141~0 14~0 I~ 013~0 141~0 '450
Holding temperature (~
(b) 1.2 S i
Fig. 9 - - ( a ) and (b) Effect of holding temperature on mechanical properties and retained austenite content at m a x i m u m U.EI and the variation
with the degree of Si and M n additions.

a similar change in U.E1 is recognized when Si content Figure 10. The product increases with retained austenite
varies. As a result of the changes mentioned above, the content until it reaches 14 pct. In addition, a comparison
product of TS and T.E1, a criterion of strength-ductility between steels A and C and between steels F and G shows
balance, is augmented by larger additions of Si or Mn that increasing Si makes the product larger, even though
and exceeds 27,500 MPa. tact. Holding at 375 ~ to 425 ~ the retained austenite content remains the same. On the
leads to these maximum values of the product, and the other hand, an increase of Mn brings an increase of re-
temperature is raised with additional Si or Mn. tained austenite content, but its relation with the product
The correlation between retained austenite content and of TS and T.E1 is not likely affected by Mn content, as
the product of TS and T.E1 (Figure 9) is summarized in shown in the bottom chart in Figure 10.

494--VOLUME 22A, FEBRUARY 1991 METALLURGICAL TRANSACTIONS A

 28000 maximum U.E1, obtainable U.E1 is not affected by Mn
1.8 Mn contents as much (Figure 11 (b)). As a result, with a fur-
26000 Steel F G H ther addition of Mn, the maximum product of TS and
S i o(y~) 0.8 1.2 1.6
450~ (I) 9 +, T.E1 generally increases according to a rise in TS. When
comparison is limited to the same holding temperature
24000 425~ O ~> 9 t / (400 ~ according to the difference in Si contents, me-
400~ [] [] 9
@A chanical properties and retained austenite contents show
22000 375~ A ~l 9 / the same pattern of change mentioned in the former sec-
350~ ~7 ~7 9 ~7
tion, even though the cooling rate varies.
20000 I I I I To understand the change in mechanical properties with
28OO0 the cooling rate, the fraction of microstructure was quan-
0.9 Mn ( ~ ~ / ! tified. An example is Figure 12, showing the case for
26000 Steel A B C steel E. If cooling is slower than the optimum rate to
S i 0(%) 1.2 1.6 2.0 enhance retained austenite content or to elevate the prod-
4500C O (D 9 uct of TS and T.E1 (about 70 ~ in this case), with
,T, 425~ O ~> 9 a decline of the cooling rate, the fraction of both bainite
400~ [] [] 9 and retained austenite becomes smaller and smaller due
x 22000 375~ A ,ix 9
09 to the growth of ferrite during cooling. A fall in TS,
F- 350~ ~7 q7 9
found in Figure 11, coincides with this increase of fer-
20000 I I
rite. In the case of smaller Mn additions or still slower
28000
;reel A D E G
Ste, cooling, pearlite appears and a remarkable drop in U.E1
Mn 0(%)
( 0.9 1.2 1.5 1.8 and T.E1 is recognized. On the other hand, faster cooling
26000 450'
450~ O (D () 9 over the optimum rate also results in a decreased content
425
425~ O <I> <~ 9 of retained austenite, although the fraction of ferrite re-
24000 4oo~
400 [] [] 9 9 b N
mains the same. Supposedly, this decrease is caused by
375~
375 Zl A A 9 ,FLI~,'
150~ ~)q7 ~ 9 .,.Y~-T~
the growth of ferrite with C in supersaturation or mar-
35O
22OO0 tensitic transformation. Though such a difference in
l microstructure is produced by a change in the cooling
A 1 I2 S i
O , - ~ . . . . . . . . Tv l I rate, the correlation between retained austenite content
2OO0O
0 4 8 12 16 20 and the product of TS and T.E1 obeys the same zone,
Retained austenite content o(fl~) as shown in Figure 10. In addition, the effect of Si and
Mn contents on the relation and inverse correlation above
Fig. 1 0 - - R e l a t i o n between retained austenite content and the product
of TS and T.E1 f o r varied holding temperatures. 14 pct of retained austenite holds true also in this case.

D. Influence of the Cooling Rate on the Resulting IV. DISCUSSION

Properties and Its Variations with Si and Mn Contents
A. Mechanism of Austenite Stabilization
In Sections III-A through I l l - C , results given were
obtained by the experiments with salt baths. Therefore, By quenching from a single austenite range, retained
the cooling rate to the bainite temperature was constant austenite is easily formed in steel with a large C content,
at around 150 ~ This section, however, is concerned and holding at the bainite transformation range for a while
with the results of varied cooling rates; the holding tem- after its interruption increases its content greatly when a
perature is fixed at 400 ~ The results are summarized large amount of Si is added to steel, t6-17] This is attrib-
in Figure 11 which presents the mechanical properties uted to the inhibition of carbide precipitation by Si which
and retained austenite content when the holding time is does not dissolve in carbide. At the same time, C is en-
set to attain the largest U.EI at a given cooling rate. Faster riched into the remaining austenite and its Ms temper-
cooling always results in a slightly higher TS. Uniform ature is greatly reduced. Some reports showed blocky
elongation and T.EI are recognized to reach the maxi- retained austenite, the size of which sometimes exceeded
mum at a certain cooling rate, depending on steel chem- 10/zm in diameter t17] but most were film-shaped in the
istry, and to become smaller and smaller, even though lath of upper bainite and were difficult to identify by
the rate increases beyond that maximum. The degree in optical microscopy measurements. However, this lath type
variations of U.E1 and T.E1 with cooling rate is larger of retained austenite is scarcely found in the steels with
than that of TS; this implies the influence of the cooling lower C and Si contents used in this study, as recognized
rate on the product of TS and T.EI resembles the influ- from Figure 3. Instead, retained austenite brought by
ence of the rate on U.E1 and T.E1. The effect of Si on intercritical annealing appears as very fine particles, as
the optimum rate is difficult to discern, as shown in presented in Figure 4(b). Its fraction estimated by opti-
Figure 1 l(a). However, in the case of lower Mn addi- cal microscopy measurements almost corresponds to that
tions, as in steel D in Figure 1 l(b), faster cooling is needed determined by the X-ray diffraction method. Therefore,
to maximize U.E1, T.E1, and the product of TS and T.E1. film-shaped retained austenite is negligible in this case.
The variation of retained austenite content with the cool- Irrespective of morphological differences, this austen-
ing rate resembles that of U.E1, T.E1, and the product ite is stabilized due to C enrichment: the same mecha-
of TS and T.E1. If the cooling rate is chosen to give the nism as in higher C steel cooled from the single austenite

METALLURGICAL TRANSACTIONS A VOLUME 22A, FEBRUARY 1991--495

 1100 401 30000
16
C(2.0 S i) LU
D
1000 351B(1.6S i ) ~ ~ 27500 X
t~
E 12
C
T.EI ~ A" A
#_ 25000
3;
v

co
900 30 /A(1 2 Si)
o
!
c,~
," B
7 E
.I-i
c-
o
o
8
b-
800 25 a , ,,,A
o 22500 .*2_
X
t--
5 ff U.EI
I! t-- 4-1
co
H
-~ 4
700 20 ,, 200013
Im

60O 10
I I I I I I I I
15 10 20 50100 17500 i I i ~ rr 0~ i i
20 50 100 10 20 50 100 10 20 50 100

(a) 0.9 Mn Cooling rate ( C / s )

1100

1000
~
35T.
~16
uJ

E]2

900 n
d~ 3; e-
3; G(1.8Mn) O
[]
i <> 4-1

iii
co
I-- 8001-o,i O ~ 25 -~
1(1.5Mn)2 "O G X 1-
co
b
I .,dDO2Mn E
7001 / 20r .;
e-
I'-' A(0"9 Mn)l
I1ID
0
600% 2b 5b1 o' I
15 b2b 50 I
100 ..... 10 20 50 100
rr
10 20 50 100
Cooling rate (~
(b) 1.2 S i
Fig. 1 1 - - ( a ) and (b) Effect o f c o o l i n g rate on m e c h a n i c a l properties and retained austenite content at m a x i m u m U.EI at 400 ~ and the variation
with the d e g r e e o f Si and M n additions.

range. Since soaking at the intercritical range is prac- creased. As the chemical composition of austenite grains
ticed after cold-rolling, extremely fine austenite enriched formed at intercritical annealing resembles that of pre-
with C is formed along ferrite grain boundaries. [27] viously investigated steels, [6,TAS,E~ similar structures
At intercritical annealing, the rim of austenite grains composed of film-shaped ferrite and austenite may be
becomes lean in Si which tends to be partitioned more formed but actually are not. One of the reasons is that
to ferrite than to austenite. When a hold at bainite trans- from a view of boundary energy, movement of the inter-
formation range starts, this rim is preferentially trans- face has an advantage compared to forming a new
formed to bainitic ferrite, rejecting C to the remaining boundary between ferrite and austenite in lath structure.
core. Then, stability of the austenite core is greatly in- Transformation of the brim of austenite to ferrite is also

4 9 6 - - V O L U M E 22A, FEBRUARY 1991 METALLURGICAL TRANSACTIONS A

 100 into austenite and is enhanced at a little higher holding
temperature by a further addition of Si. In short, further
alloying of Si improves the stability of retained austenite
and results in superior strength-ductility balance.
80 Figure 10 also indicates the product of TS and T.E1
begins to decrease even if the retained austenite content
continues to increase beyond 14 pct. This corresponds
60 to the saturation of U.EI beyond 1.5Mn (Figure 9(b))
t- and is caused by the difference of C content in retained
O
austenite. The C content in steel, as a whole, is 0.2 pct;
0
40 Ferrite so the C content in retained austenite changes with its
63
amount, as indicated by the dashed line in Figure 13, on
LL
the condition that ferrite and bainite contain no carbon
and that carbide is not precipitated. However, predicted
20
from the free energy analysis and verified experimen-
tally, transformation to bainite starts if C in austenite
surpasses 1.7 pct at around 400 ~ t8] This means the
0 ~ • change should be like the continuous line in Figure 13,
10 20 50 100
and the retained austenite with the highest stability can-
Cooling rate ( ~ not be higher than 12 pct in steel of 0.2C. Retained aus-
Fig. 1 2 - - C h a n g e s in fraction of microstructure with cooling rate
tenite contents greater than 12 pct will have lower C,
(steel E). although differences of 12 and 14 pct arise from the
uneven C distribution. Therefore, the quality and quan-
tity of retained austenite vary with Si and Mn contents
and holding temperature within a limit determined by the
preferable in releasing strain brought about by the greater
C content of steel.
shrinkage of austenite than that of ferrite.

B. Enhanced Elongation by Retained Austenite and Its C. Determination of Optimum Alloying Considering
Dependence on Alloying and Carbon Content the Influence of Cooling Rate
Properties of dual-phase steel are strongly dependent
There has been a discrepancy as to which type of re-
on its Mn content, because Mn strongly affects its
tained austenite is effective in improving tensile ductil-
hardenability; so when the effect of Mn is discussed, the
ity. Some reports 128'29]concluded that the film-shaped one
effect of the cooling rate must be considered at the same
in bainite lath is effective but that the granular one is
time. This holds true also for the present study. How-
not. Contrarily, others recognized the benefit of granular
ever, even though a comparison is made at optimum
ones.t3.4] This study concerns fine grains of retained aus-
cooling rates according to Mn content, TS and the prod-
tenite which bring about TRIP remarkably, as shown in
uct of TS and T.E1 increase with the addition of Mn, as
Figure 6.* There may be reasons for the discrepancy,
*The retained austenite is, however, stable to liquid nitrogen
temperatures. 2.5 \
among which the difference of C enrichment is essential. %
Its small size, less than about 1 /zm in diameter, is also
important compared with previously reported retained r 2.0 - %
austenite ranges of about 4 to 5 / x m I2~ and sometimes %
%
beyond 10/zm. tim Coexisting pure and highly deform- ||l i
63
able ferrite typical of intercritical annealing may also be
important, because the concentration of strain can be "o
1.5-
r
avoided compared with formerly studied structures. , B

63
As mentioned in reference to Figure 10, steel with a
larger Si content has the larger product of TS and T.E1 1.0-
despite a slight change in retained austenite content. This r - - - a s s u m i n g that all C is contained
is attributed to its stability difference, as shown in in retained a u s t e n i t e
Figure 6. In steel C, containing a larger amount of Si, t-
strain-induced transformation occurs gradually and so 0.5 - _ _ assuming 'that retained austenite
t-
larger U.E1 is brought on. The stability corresponds to O cannot exist w i t h i t s C content
the C enrichment. At the start of bainite transformation, 0
beyond 1.7 %
C is unevenly distributed in austenite, but sufficient o 0 I I I )
holding enriches the remaining austenite uniformly with 0 4 8 12 16 20
C to the maximum. Figure 7 shows that by increasing
Si content, time to attain the largest U.E1 is prolonged Retained a u s t e n i t e c o n t e n t ~
more than time to reach the.top retained austenite con- Fig. 1 3 - - S u p p o s i t i o n of the m a x i m u m C content in retained austen-
tent. This gap corresponds with sufficient moving of C ite for its different amounts.

METALLURGICAL TRANSACTIONS A VOLUME 22A, FEBRUARY 1991--497

mentioned in Section III-D. Also, the correlation be- REFERENCES
tween retained austenite content and the product of TS
and T.EI and its variation with Si and Mn contents are 1. S. Hayami and T. Furukawa: Microalloying 75, Union Carbide
Corp., New York, NY, 1975, pp. 78-87.
not changed. So, of course, choosing the moderate cool- 2. T. Furukawa, H. Morikawa, H. Takechi, and K. Koyama: Structure
ing rate determined by steel chemistry is important in and Properties of Dual-Phase Steels, R.A. Kot and J.W. Morris,
order to get better properties, but even more important eds,, AIME, New York, NY, 1979, pp. 281-303.
is determining the amounts of alloying elements and the 3. Narasimha-Rao V. Bangaru and A.K. Sachdev: Metall. Trans. A,
1982, vol. 13A, pp. 1899-906.
holding condition following the cooling, as the resulting 4. A.K. Sachdev: Acta Metall., 1983, vol. 31, pp. 2037-42.
properties strongly depend on them. These facts being 5. V.F. Zackay, E.R. Parker, D. Fahr, and R. Bush: Trans. Am.
considered, steel C (2.0Si-0.9Mn) may be selected for Soc. Met., 1967, vol. 60, pp. 252-59.
rapid cooling and steel G (1.2Si-l.8Mn) for slow cool- 6. S.J. Mates and R.F. Hehemann: TMS-AIME, 1961, vol. 221,
pp. 179-85.
ing. A desirable holding temperature is 425 ~ for both
7. R. Le Houllier, G. Begin, and A. Dube: MetaU. Trans., 1971,
steels, but some modification will be necessary to meet vol. 2, pp. 2645-53.
line requirements for production on an existing contin- 8. V.M. Pivovarov, I.A. Tananko, and A.A. Levchenko: Phys. Met.
uous annealing line. Metallogr., 1972, vol. 32, pp. 116-22.
9. E. Dorazil and J. Svejcar: Arch. Eisenhiittenwes., 1979, vol. 50,
V. CONCLUSIONS pp. 293-98.
10. R.F. Hehemann, K.R. Kinsman, and H.I. Aaronson: Metall.
Silicon and Mn containing sheet steel of simple chem- Trans., 1972, vol. 3, pp. 1077-94.
11. H.K.D.H. Bhadeshia and D.V. Edmonds: Metall. Trans. A, 1979,
istry have been processed by intercritical annealing fol-
vol. 10A, pp. 895-907.
lowed by bainite transformation to produce large quantities 12. H.K.D.H. Bhadeshia and D.V. Edmonds: Met. Sci., 1983, vol. 17,
of retained austenite. This retained austenite is very fine pp. 411-19.
and stabilized by C enrichment. Transformation of re- 13. H.K.D.H. Bhadeshia and D.V. Edmonds: Met. Sci., 1983, vol. 17,
tained austenite to martensite results in high strength and pp. 420-25.
14. H.K.D.H. Bhadeshia and D.V. Edmonds: Acta Metall., 1980,
ductility in these steels. The product of TS and T.E1 ex- vol. 28, pp. 1265-73.
ceeds 27,500 MPa.pct for steel with 2.0Si-0.9Mn or 15. H. Seilstorfer and N.B. Ottobrunn: Arch Eisenhiittenwes., 1974,
1.2Si-l.8Mn held at 425 ~ Other findings are de- vol. 45, pp. 623-27.
scribed below. 16. B.P.J. Sandvik and H.P. Nevalanien: Met. Technol., 1981, vol. 8,
pp. 213-20.
1. Longer holding in the bainite transformation range is 17. V.T.T. Miihkinen and D.V. Edmonds: Mater. Sci. Technol., 1987,
required to get the maximum U.E1 with increasing Si vol. 3, pp. 422-31.
and Mn. 18. V.T.T. Miihkinen and D.V. Edmonds: Mater. Sci. Technol., 1987,
vo|. 3, pp. 432-40.
2. The optimum cooling rate to get retained austenite is 19. V.T.T. Miihkinen and D.V. Edmonds: Mater. Sci. Technol., 1987,
reduced and the optimum holding temperature is raised vol. 3, pp. 441-49.
by increasing the Mn addition. Stability of the ob- 20. K. Tomita, T. Okita, and K. Nakaoka: Trans. lron Steel Inst.
tained retained austenite does not change, but the Jpn., 1984, vol. 24, p. B363.
possible content brought about by the optimum pro- 21. O. Matsumura, Y. Sakuma, and H. Takechi: Trans. Iron Steel
Inst. Jpn., 1987, vol. 27, pp. 570-79.
cess is increased with more additions of Mn. 22. O. Matsumura, Y. Sakuma, and H. Takechi: Scripta Metall., 1987,
3. Additional amounts of Si do not affect the desirable vol. 21, pp. 1301-06.
cooling rate to get retained austenite and its resulting 23. H.C. Chen, K. Tomokiyo, H. Era, and M. Shimizu: Tech. Rep.
amount, but an increase of Si improves the stability Kyushu Univ., 1987, vol. 60, pp. 643-49.
24. K. Tsuzaki, J. Inada, K. Matsumoto, and M. Maki: Current Adv.
of retained austenite and makes U.EI and the product in Mat. and Processes, 1988, vol. 1, p. 1855.
of TS and T.E1 larger. This enhancement is dominant 25. K.W. Andrews: J. Iron Steel Inst., 1965, vol. 203, pp. 721-27.
at higher holding temperatures. 26. S. Bandoh, O. Matsumura, and Y. Sakuma: Trans. Iron Steel
4. Stability of retained austenite depends chiefly on its Inst. Jpn., 1988, vol. 28, pp. 569-74.
C content. An excessive addition of Si or Mn in- 27. D.Z. Yang, E.L. Brown, D.K. Matlock, and G. Krauss: Metall.
Trans. A, 1985, vol. 16A, pp. 1385-92.
creases the retained austenite content but does not im- 28. K. Shinoda and T. Yamada: Netsu Shori, 1980, vol. 20,
prove the combination of strength and ductility, since pp. 326-30.
this retained austenite has lower overall C and, there- 29. S. Tagashira, T. Yamada, T. Kamiyo, and K. Shinoda: Curr.
fore, lower stability. Adv. Mater. Processes, 1988, vol. 1, p. 1774.

498--VOLUME 22A, FEBRUARY 1991 METALLURGICAL TRANSACTIONS A