Usual, normal thermo- mechanical processing of SDSS F55 alloy [611795]

Chapter 2
Usual, normal thermo- mechanical processing of SDSS F55 alloy

2.1 Introduction
Recently a duplex stainless steel, which has a two -phase ferrite ( alpha ) + austenite ( gama )
microstructure, has attracted great interest with its cost- saving combination of high strength and
improved resistance to general and localized corrosion, stress -corrosion cracking, abrasion and wear
[18, 19] . In case of two -phase alloys, a refinement of matrix grains and a uni form distribution of fine
second phases are essential in order to improve mechanical properties.
Thermo mechanical processing, which combines defor mation and heat treatment, is very effective for
microstruc ture control and hence for the improvement of mechanical properties of metallic materials.
For example, in the case of Fe- Cr-Ni duplex stainless steel, a micro duplex structure consisting of
Alfa and Gama in a fine -grained form (1 -3 pm) is obtained by the proper thermo mechanical
processing. The micro duplex structure exhibits a super plasticity at elevated temperatures [20-25,
18], and a high strength, a high fatigue strength and a good toughness at room temperature [26, 27,
18]. In the pre sent paper, the change of two- phase microstructure, espe cially micro duplex structure,
by various types of thermo mechanical processing and heat treatments is briefly reviewed mainly
based on our experimental results.
2.2 Formation Process of Micro duplex Structure by Thermo mechanical Processing
Duplex stainless steels contain high amounts of Cr and Ni. One of the typical chemical compositions
is Fe – 25%Cr -7%Ni -3%Mo. In this steel, the matrix phase is bcc ferrite ( a) and the second phase is
FCC austenite ( g) which is precipitated from a matrix. The volume fraction of g is about 40- 45%.
Fig. 2. 1 shows a calculated phase diagram of Fe- Cr-7%Ni -3%Mo alloy. When the duplex stainless
steel is cooled slowly from a single -phase region, very coarse two- phase structure is formed (Fig. 2.2
a). on the other hand, when quenched from a single -phase region an d then aged at temperatures in
the two -phase region Fig. 2.1 (a), g phases precipitate along coarse a boundaries and within a grains
as shown in Fig . 2.2 (b) [28] . This two- phase structure obtained by quenching and aging is finer than
that of slowly cooled specimen. However, the distribution of g phase is not so uniform and g size is
still fairly large. Furthermore, a matrix grains are very huge since the a grain size is determined by
the solution- treatment at higher temperatures.
In order to obtain much fi ner two -phase structure, thermo mechanical processing is necessary. The
typical thermo mechanical processing in the duplex stainless steel is shown in Fig . 2.1. (b).The

specimen is solution -treated at higher tem peratures in a single -phase region and water -quenched to
obtain a supersaturated a at room temperature. The solu tion-treated specimen is heavily cold -rolled
(70-90 %) and then aged at temperatures around 1 273K in a+g region. With this treatment, very fine
g particles are uniformly precipitated as shown in Fig . 2.2 (c) [29]. This structure is the micro duplex
structure. However, this micro duplex structure is not always obtaine d by the treatment shown in Fig.
2.1 (b). In general, when the supersaturated matrix is aged after heavy cold rolling, two metallurgical
processes, i.e., a recovery or a recrystallization of matrix phase and a precipitation of second phase,
occur simultaneously during aging. By the competition of these two processes, final microstructure
after aging changes in various ways depending on the amount of cold rolling and aging temperature.

In the case of duplex stainless steel, three types of a+g structures are obtained by the aging after
heavy cold rolling of supersaturated a as shown in Fig . 2.3 [29] , in which Fig. 2.3(a)

Is the micro duplex structure consisting of fine a and g. Fig . 2.4 [29] summarizes a formation process
of three types of a+g structure in connection with RPTT (Recrys tallization -Precipitation –
Temperature- Time) diagram. It is characteristic that the recovery of deformed a occurs rapid ly and
the sub grain structure of a matrix is formed readily prior to the precipitation of g phase in the earliest
stage of aging. At low er aging temperatures Fig.2. 4 (a) and 2.4(b), g phase precipitates at a sub grain
boundaries as shown in Fig. 2.5 [29]. By further hol ding at lower temperatures Fig. 2.4 (a), the volume
fraction of g is increased to about 0.4 and the micro duplex structure consisting of fine a sub grain
matrix and uniformly distributed fine g precipitates is obtained Fig. 2.3(a). On the other hand, when
the specimen is a ged at higher temperatures Fig.2. 4 (b), the volume fraction of g becomes small and
hence the recrystalli zation of a sub grain matrix occurs by prolonged aging, because a pinning effect
of g particles on the migration of a sub grain boundaries is reduced. These results in the coarse two –
phase struc ture consisting of large recrystallized a grains and fairly large g precipitates as shown in
Fig. 2.3(b). Furthermore, when the specimen is aged a t much higher temperatures Fig. 2.4(c), the
recrystallization of a matrix occurs prior to the precipitation of g, resulting in coarse a grains and film
like g precipitate s formed along a grain boundaries Fig. 2.3 (c).
2.3 Nature of Micro duplex Structure in Duplex Stainless Steel s
Fig. 2.6 (a) [30] is a typical TEM micrograph of micro duplex structure in Fe -26%Cr -8%Ni alloy aged
at 1 273K for 60 s after 85 % cold rolling of supersaturated a. The aver age sizes of a sub grains and
g particles are 1.2 mm and 1.0 mm, respectively. Fig . 2.6 (b) shows a distribution of mis – orientation

angle between a grains measured by Kikuchi pattern analysis. It appears that most of a grain
boundaries are of low angle type with a miss orientation of less than 5°, i.e., sub grain boundaries.
Figure 2.6(c) is a histogram of mis s orientation from K -S relationship between g and adjacent a sub
grain matrix. A deviation from K -S relationship is also small (less than 10°).
Therefore, it can be concluded that the micro duplex structure such as Fig . 2.2 (c) and Fig. 2.6(a)
is characterized by the a sub grain matrix and fine g particles precipitated pre dominantly at a sub
grain boundaries. Since the a matrix of micro duplex structure is a recovered structure, the micro
duplex structure has a strong texture ({001}(110) and (111}(110) a texture) which is the same as cold
rolling tex ture of a. Fig . 2.7 [31] shows the micro texture of a sub grains in micro duplex structure,
displayed as (a) {100} and {111} pole figures and (b) ND and RD inverse pole figures, where ND is
the normal direction of the rolling plane and RD is the rolling direction of the specimen. It is clear
that orientations of a sub grains concentrate to the (001)[110] and (111)[110] orientations. It was
confirmed that there is no significant difference in th e miss orientation distribution of a sub grains in
two regions with different textures [31].
In the case of (g+ a) two -phase alloy with the g matrix such as Ni -Cr alloy, the formation process of
micro duplex structure is entirely different from the a matrix alloy like duplex stainless steel [32]. In
the fcc g matrix alloy, the recov ery of deformed matrix hardly proceeds, and then precipita tion of a
phase occurs in the deformed g containing a high density of dislocations. After precipitation of a, the
g matrix undergoes recrystallization, resulting in the micro duplex structure consisting of fine
recrystallized g grains and finely precipitated a particles. Therefore, the nature of a+g micro duplex
structure formed by thermo mechanical processing is different in the a matrix alloy and the g matrix
alloy, because of the difference in the extent of recovery in the matrix phase [32].

2.4 Microstructure change of m icro duplex s tructure du ring p rolonged aging
Fig 2.8 shows optical micrographs of the specimens aged at 1 273K for (a) 1.8 ks and (b) 360 ks after
80% cold rolling of supersaturated a in Fe -26%Cr -7%Ni alloy. Initial micro duplex structure Fig.
2.8(a) exhibits a significant coarsening of g particles by the lo ng time holding for 360 ks Fig 2.8(b),
whereas the volume fraction of g phase is independent of holding time and is 0.36. The a/a boundaries
are hardly etched and invisible in t he optical microstructure Figure 2.8(b) even after the long time
aging. This indicates that a/a boundaries are still low angle bound aries, i.e., sub grain boundaries. In
the case of aging for 1.8 ks Fig. 2.8(a) sizes of a sub grains and g particles are 2.6 mm and 2.2 mm,
respectively. They r each 15.6 mm a nd 13.8 mm by 360 ks aging Fig. 2.8(b). As is shown in Fig. 2.9
[31], the cube of grain size versus aging time satisfies a linear relationship for both a sub grains and

g particles. This indicates that the grain growth of both phases in micr o duplex structure satisfies the
third power law as has been reported by Abe et al[33]. The sub grain size of a matrix is controlled by
the size of g particles following Zener’s relation, i.e., d=p(r/f ), where d is the size of a sub grain, r
and f are the size and volume frac tion of g particles, respectively, and p is a constant. The co efficient
p in Zener’s relation is about 0.42 in the present case. Fig. 2.10 [31] shows the distribution of mis s
orientation angle between adjacent a sub grains in micro duplex struc ture for three aging conditions
at 1273K. The average mis s-orientation angle hardly changes even after the a sub grain si ze increases
up to about 16 mm. This insignificance in the mis s orientation with sub grain growth res ults from the
fact that the variation of local lattice rotation in the a matrix is not monotonic but periodic, and no
long range lattice cur vature exist [31].
Therefore, it can be concluded that the a matrix in micro duplex structure does not recrystalliz e, and
maintains the sub grain structure even after the prolonged holding at aging temperature because sub
grain boundaries are pinned by g particles.

2.5 Microstructure Change of Micro duplex Structure by Annealing after Heavy Cold Rolling
Cooke et al}n) studied the microstructure change of micro duplex structure during annealing at 1 173K
after 40% cold rolling in Fe -28%Cr- 11%Ni alloy, micro duplex struc ture where the volume fraction
of g phase is 0.45 and the sizes of a sub grains and g particles are a few microns. They showed that
the a matrix does not recrystallize whereas the recrystallization takes place in g phase. They did not,
however, reported whether the a matrix is still hard to recrystallize in case of heavy cold rolling over
40% in reduction. On the other hand, Blicharski [35] studied the re crystallization behavior of coarse
a+g structure in Fe – 26%Cr- 7%Ni alloy, where g particle size and the mean free distance of a are
about 100 mm, and reported that recrystal lization of a matrix occurs in the 43% cold- rolled speci men.
These studies suggest that the occurrence of recrystal lization of a matrix in a+g structure depends on
the size of initial micro structure in the case of about 40% cold – rolled specimen.
Huang et al. [36] studied a possibil ity of recrystallization of a matrix in the heavily cold -rolled (80
%) micro duplex structure with different sizes of a and g (from about 2 mm to about 16 mm) in Fe –
26%Cr -7%Ni alloy. Fig . 2.11 [36] is an exam ple of optical micrograph of the specimen which was
annealed at 1 273K for 4.2 ks after 80% cold rolling of micro duplex structure shown i n Fig. 2.8 (a)
(sizes of a and g are about 2.5 mm). The a matrix is smooth and no grain boundary is seen, indicating
that the a matrix is a sub grain struc ture and the recrystallization does not occur in a matrix. Fig . 2.12
[35] shows ND and RD inverse pole figures for a sub grains determined by Kikuchi pattern analysis
in the specimen of Fig . 2.11. The strong texture component of (111}(110) is observed. The average
miss orientation angle between a sub grains is 5.4°, which is not significantly dif ferent from that in
the starting materials, 4.2° Fig . 2.10 (a). It was confirmed that the recrystallization of a matrix does

not occur when the initial structure before rolling is fine, i.e., the a sub grain size is smaller than about
10 mm. On the other hand, in the case of coarse initial structure in which the size of a sub grain is
about 16 mm, the recrystallization occurs in the a matrix with the (111}(110) initial orienta tion. The
{001}< 110) initially oriented region, by contrast, does not recrystallize even in the coarse initial
structure. This result indicates that the a matrix in micro duplex structure with fine a and g (less than
about 10 mm) is very hard to recrystallize even by the heavy cold rolling (80%) and annealing.

2.6 Microstructure Change of Micro dupl ex Structure during Superplastic Deformation
The duplex stainless steel with micro duplex structure ex hibits a super plasticity at elevated
temperatures [3-8] although the a matrix is sub grain structure, which is not favorable to grain
boundary sliding. In order to make clear the reason why the micro duplex structure shows a large
elongation at high temperature s, W.SCUMACKER and L.COUDREUSE [8,13] observed the
microstructure change during superplastic deformation in micro duplex structure of Fe -25%Cr -7%Ni –
3%Mo alloy. Fig . 2.13 [25]. shows TEM micrographs of the specimen deformed by 20% and 85% at
1 273K at a str ain rate of 1.7X10“2s_1. In the specimen deformed by 20% Fig . 2.13 (a), the dislocation
density is fairly high in both phases of a and g, indicating that the deformation takes place by the slip.
On the other hand, when the specimen is deformed up to 85%, the microstructure changes drastically as shown in Fig . 2.13 (b). Dislocations are scarcely observed in both phases and the mis s orientation
between adjacent a grains is large. It was confirmed that the dislocation density was very low any –
where in the sp ecimen deformed by 85 -1 300%, indicating that the deformation takes place by grain
boundary sliding.
Fig. 2.14 [30] shows a change in the mi ss orientation be tween a sub grains (a) and the mis s
orientation from K -S relationship between g and adjacent a matrix (b) during the early deformation
up to 100% elongation of micro duplex structure in Fe -26%Cr -8%Ni alloy. It appears that the mis s-
orientation of a sub grains gradually increases with an in crease in the amount of deformation and
most of a boundarie s become high- angle boundaries after 100% elongation. The mis s orientation
from K -S relationship between a matrix and g particle is also gradually increased as the tensile
deformation proceeds.
Fig. 2.15 [25] is a summary of observations of the mi crostructure change during superplastic
deformation of micro duplex structure at 1 273K at a strain rate of 1.7X10-2 s_1 in Fe -25%Cr -7%Ni –
3%Mo alloy. Before deformation, ala boundaries are of low -angle (sub grain boundaries), and alg
interphase boundaries are coherent. In the early stage of deformation (20 -50% elongation), the
deformation takes place by the usual slip. During this stage, a sub grain mis s orientation is gradually

increased with an increase in the deformation, and finally a sub grains change to grains with high –
angle boundaries. This phenomenon is a dynamic continuous recrystallization, and it is completed at
about 85% elongation in the present case. Once the fine and equiaxed micro duplex structure with
high-angle boundaries is formed, the grain boundary sliding becomes a dominant de formation mode
and the super plasticity appears.
Therefore, it is concluded that the role of dynamic con tinuous recrystallization in the superplastic
deformation is to make the fine two -phase structure suitable for the grain boundary sliding in the early
stage of deformation. Similar microstructure change during superplastic deformation has been
reported in aluminum alloys [20-23]. The gradual increase in the mis s orientation of a sub grains in
the early stage o f deformation is accounted for by the repeated ab sorption of dislocations, which had
been introduced to ac commodate the plastic incompatibility of neighboring a and g phases
undergoing slip deformation, into sub grain boundaries.[25,41]
2.7 Microstructur e Change of Micro duplex Structure by Cyclic Treatment of Slight Cold
Rolling and Annealing
Micro duplex structure in the duplex stainless steel consists of fine g particles in the a sub grain
matrix. If sub grains of a matrix become fine grains with high -angle boundaries, more excellent
mechanical properties at room temperature can be expected. However, as described above, a sub
grains in micro duplex structure are very stable even after the long holding time at aging temperatures
and the recrystallization of a sub grains matrix hardly occurs even by the heavy cold rolling and
annealing. Based on the information obtained by the study on dy namic continuous recrystallization
during superplastic de formation [25], Tsuzaki, and Maki [41] tried to convert a sub grains in micro
duplex structure into a grains with high -angle boundaries by the repetition of cycles of slight cold
rolling (10%) and short time annealing (1 273K, 0.3 ks). In this cyclic treat ment, we aimed the
repetition of introduction of disloca tions in a sub grains by the slight cold rolling and the absorption
of these dislocations into sub grain boundaries by the short time annealing. Fig . 2.16 [41] shows the
distribu tion of mis s orientation angle between adj acent a sub grains for the starting materials (a) and
the specimens after 3 cy cles (b) and 14 cycles (c) of the repetition. With an increase in the number of
cyclic heat treatments, the fraction of high -angle boundaries, across which mis s orientation is more
than 15°, is increased. The average mis s orientation of a sub grains is increased continuously from
4.2° (initial structure) to 10.3° (after 14 cycles). This is a new thermo mechanical processing which
converts sub grains into grains with high- angle boundaries in the micro duplex structure.
2.8 Formation of Micro duplex Structure by Annealing after Heavy Cold Rolling of Coarse
(a+g) Structure

As was described in Fig . 2.1, micro duplex structure in the duplex stainless steel is usually produced
by the thermo mechanical processing which involves the following process es; (1) solution treatment
at higher temperatures in a single-phase region, (2) quenching to obtain a supersaturated a single –
phase, (3) heavy cold rolling of supersaturated a, and (4) aging at a+g region. Instead of this
complicated treatment, Furuhara, and Maki [42] tried to obtain the micro duplex structure by more
simple treatment that is just annealing after heavy cold rolling of hot- forged bars with an elongated
coarse two -phase struct ure. As a starting material, we used hot -forged bars in Fe -26%Cr -8%Ni alloy
as shown in Fig . 2.17 (a) [42] . Both a and g phases are elongated along the longitudinal direction of
hot-forged bar. Fig . 2.17(b) [42] is a three- dimensional SEM micrograph of the specimen annealed
at 1 273K for 60s after 85% cold rolling of hot -forged bar. Elongated g layer in the cold -rolled
specimen turns to a row of equiaxed g grains. A disconnection of g grain alignment occurs to some
extent by annealing. This microstructure is similar to the micro duplex structure obtained by the
conventional thermo mechanical processing shown in Fig . 2.2(c). TEM observation indicated that the
micro duplex structure of Fig . 2.17(b) c onsists of a sub grains and recrystallized g grains. It has been
reported that the hot -rolled or cold- rolled duplex stainless steels with elongated a and g grains exhibit
a superplastic deformation at elevated temperatures without any further heat treatments [26, 27˄] .
Furuhara, and Maki [42] also confirmed that the structure like Fig . 2.17(b) exhibits a large
superplastic elongation nearly the same as the micro duplex structure produced by the conventional
thermo mechanical process ing. For the appearance of super plasticity in the duplex stainles s steel, it
has long been believed that the complicated thermo mechanical processing such as Fig . 2.18(a) is
neces sary to obtain the micro duplex structure. However, it appears that the micro duplex structure
and then the super plasticity can be obtained by the simple treatment without solution treatment as
shown in Fig . 2.18(b).