See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/226115961

Development of microstructural banding in low-alloy steel with simulated Mn

segregation

Article in Metallurgical and Materials Transactions A · January 2002

DOI: 10.1007/s11661-002-0172-8

CITATIONS READS
102 2,992

3 authors:

Ted Majka David Matlock

General Electric Colorado School of Mines
2 PUBLICATIONS 102 CITATIONS 373 PUBLICATIONS 12,458 CITATIONS

SEE PROFILE SEE PROFILE

George Krauss
Colorado School of Mines
228 PUBLICATIONS 8,420 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

High tempertaure tempering of martenstic microstructures in carbon steels View project

Surface hardening for wear and abrasion resistance View project

All content following this page was uploaded by Ted Majka on 06 August 2018.

The user has requested enhancement of the downloaded file.

Development of Microstructural Banding in Low-Alloy Steel
with Simulated Mn Segregation
TED F. MAJKA, DAVID K. MATLOCK, and GEORGE KRAUSS

The development of microstructural banding in low-alloy steel with Mn segregation has been investi-
gated through the use of artificially segregated steel, interrupted cooling techniques, and optical
microscopy. Mn segregation was simulated by hot roll bonding thin sheets of 5140 steel with 0.82
wt pct Mn and modified 5140M with 1.83 wt pct Mn into a plate with 20- and 160-␮m-thick segregated
layers. Samples were austenitized at 850 ⬚C, continuously cooled at 1 ⬚C/s and 0.1 ⬚C/s, and quenched
from progressively lower temperatures to observe the evolution of the microstructure. The segregated
band thickness had a striking effect on microstructural development. Samples with 160 ␮m bands
cooled at 1 ⬚C/s had martensite and bainite in high-Mn bands. In contrast, samples with 20 ␮m bands
cooled at the same rate had pearlite in high-Mn bands. The dramatic effect of band thickness on
microstructural development was due to growth of a fully pearlitic band at the interface between
segregated layers. The formation of interfacial pearlite is discussed relative to redistribution of carbon
between adjacent high- and low-Mn bands during cooling.

I. INTRODUCTION wt pct Mn steel and concluded that the intensity of micro-

structural banding increases as the cooling rate decreases.
MICROSTRUCTURAL banding in low-alloy steel is The study showed that furnace cooling produces intense
due to the segregation of substitutional alloying elements
banding and air cooling yields poorly defined ferrite and
during dendritic solidification.[1–7] The addition of elements
such as manganese, chromium, and molybdenum causes pearlite bands. Increasing the cooling rate from the austenitic
condition reduces the intensity of banding because it reduces
solidification to occur over a range of temperatures and
compositions. Consequently, the dendrite cores solidify as the Ar3 temperature differences of the segregated bands. The
authors also demonstrated that ferrite-pearlite banding is
relatively pure metal while the interdendritic spaces become
completely eliminated at cooling rates greater than 5 ⬚C/s.
enriched in solute. These high- and low-solute regions are
Ferrite-pearlite banding also disappears when the austenite
elongated into parallel bands during rolling and forming
operations.[8–11] Differences in austenite transformation grain size exceeds the spacing of the segregated bands by
behavior between bands lead to the formation of a laminated a factor of 2 or 3.[14,20] In this case, the influence of the
microstructure with discrete layers of martensite, bainite, grain boundary as a preferred site for nucleation dominates
ferrite, and pearlite.[11–17] the effect of the compositional gradients. However, Samu-
Several investigations have shown Mn to be the alloying als[20] showed that banding eliminated by fast cooling or
element most responsible for the development of microstruc- large austenite grains can reappear if the material is subse-
tural banding in low-alloy steels. Voldrich[18] showed that quently reaustenitized and slowly cooled.
carbon migrates from low- to high-Mn regions during cool- Permanent elimination of microstructural banding can
ing. Kirkaldy et al.[19] explained that this redistribution of only be achieved by high temperature homogenization treat-
carbon is due to the effect of substitutional alloying elements ments that remove the underlying compositional gradi-
on the temperature at which austenite begins to transform ents.[12] However, such treatments are often beyond the range
to ferrite, i.e., the Ar3 temperature. During cooling, ferrite of economic feasibility due to the sluggish diffusion rates
forms in bands with a high Ar3 temperature and rejects carbon of substitutional elements. For example, Leslie[17] demon-
into the austenite of adjacent low-Ar3 bands, resulting in the strated that a 26 hour cycle at 1200 ⬚C would be required
formation of carbon-rich and carbon-depleted layers. Gross- to reach 90% homogenization for a plain carbon steel with
terlinden et al.[13] used an electron microprobe analyzer to Mn segregation. Grange[11] demonstrated that a similar effect
show that regions of carbon enrichment correspond to high- can be achieved by heating for 10 minutes at 1315 ⬚C to 1345
Mn locations. ⬚C and air cooling. Microstructural banding only reappeared
As stated earlier, the primary cause of banding is substitu- after deformation was introduced to the specimens prior
tional element concentration gradients. However, cooling to reaustenitizing. The reason for the effectiveness of this
rate, austenite grain size, and austenitizing temperature also treatment was not clear to Grange, since it was unlikely
influence the severity of microstructural banding. Thompson that sufficient diffusion occurred to remove the underlying
and Howell[14] investigated banding in 0.15 wt pct C, 1.40 segregation pattern.
Grange[11] demonstrated that there are aspects about
microstructural banding that are still not understood. Conse-
TED F. MAJKA, Metallurgy Manager, is with Colfor Manufacturing, quently, this study was designed to provide further insight
Inc., Minerva, OH 44657. Contact e-mail: [email protected] DAVID K. into the development of banded microstructures in steel with
MATLOCK, Armco Foundation Fogarty Professor, and GEORGE KRAUSS,
University Emeritus Professor, are with the Advanced Steel Processing and
Mn segregation through the use of artificially segregated
Products Research Center, Colorado School of Mines, Golden, CO 80401. steel, interrupted cooling techniques, and optical micros-
Manuscript submitted October 22, 2001. copy. An SAE 5140 steel with simulated Mn segregation

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 33A, JUNE 2002—1627

 Table I. Steel Compositions (Weight Percent) of SAE 5140 and SAE 5140M
Grade C Mn P S Si Cu Ni Cr Mo Al
5140 0.39 0.82 0.014 0.022 0.22 0.16 0.17 0.81 0.04 0.030
5140M 0.41 1.83 0.015 0.024 0.22 0.16 0.17 0.81 0.04 0.029

(a) (b)
Fig. 1—Eutectoid regions of the equilibrium phase diagrams for 51xx steel with (a) 0.82 wt pct Mn and (b) 1.83 wt pct Mn showing the effect of Mn on
Ae3 temperature and eutectoid carbon content.[21]

was produced by hot roll bonding thin sheets of two different 37 pct HCl solution at 70 ⬚C. The pickled strips were cold
alloys into a plate with 20- and 160-␮m-thick segregated rolled to a thickness of 0.5 mm, sheared into 125 ⫻ 90
layers. Individual samples were austenitized, cooled at spe- mm coupons, and cleaned with ethanol and a commercial
cific rates, and quenched at progressively lower temperatures abrasive cleaning pad.
to “freeze” the microstructure as it evolved. Optical micro- Figure 2 shows the procedure for fabricating an artifi-
graphs show, in a stop-motion manner, the growth of micro- cially banded plate from individual sheets of low- and high-
structural constituents during continuous cooling. This Mn steel. Sheets of the two alloys were alternately stacked
article includes a description of the procedure for fabricating into 100-sheet compacts. The compacts were gas thermal-
steel plate with simulated segregation, a review of the inter- arc welded inside 18-gage 304 stainless steel boxes. The
rupted cooling techniques, the microstructural results, and sealed boxes were placed in a furnace at 1150 ⬚C for 1
a discussion of the effect of compositional gradients on hour and hot rolled to produce a solid plate with 20- and
microstructural development. 160-␮m-thick segregated layers. Boxes were connected to
an active vacuum system during heating and rolling to
prevent oxidation of the sheets. Figure 3 shows a 100-sheet
II. EXPERIMENTAL PROCEDURE compact and a stainless steel box. Since an equal number
Laboratory heats of SAE 5140 steel with 0.82 wt pct of sheets were used to create the compacts, the resulting
and 1.83 wt pct Mn, respectively, were vacuum induction artificially banded plates have the same volume fraction
melted, cast into 50 kg ingots, and rolled into plate form. (0.5) of low- and high-Mn steel, regardless of band
The compositions of the alloys are listed in Table I. The thickness.
1.83 wt pct Mn alloy was designated as SAE 5140M (modi- Figure 4 shows the process for “freezing” the microstruc-
fied), due to a higher Mn content than that permitted for ture on a time and temperature plot. Samples of the artifi-
standard SAE 5140 steel. Figure 1 shows calculated equilib- cially banded plate, approximately 12 ⫻ 6 ⫻ 6 mm, were
rium-phase diagrams for the two alloys.[21] Note that the austenitized at 850 ⬚C for 20 minutes in an argon-atmos-
addition of 1 wt pct Mn lowers the Ae3 temperature from phere furnace and cooled at 1 ⬚C/s and 0.1 ⬚C/s. These rates
770 ⬚C to 745 ⬚C for a 0.40 wt pct C alloy and lowers the are typical for air and furnace cooling processes, respec-
eutectoid carbon content from 0.60 to 0.56 wt pct. For tively. Individual samples were quenched in water at 10 ⬚C
convenience, the two grades of SAE 5140 will be referred intervals from 750 ⬚C to 600 ⬚C for both cooling rates,
to as low- and high-Mn for the remainder of this article. to change the untransformed austenite to martensite and
The plate was hot rolled into a 2-mm-thick strip with a capture ferrite and pearlite growth in situ. Samples were
laboratory rolling mill. The scale and decarburization cre- also quenched at 50 ⬚C intervals from 600 ⬚C to room
ated during hot rolling were removed by pickling with a temperature for the 1 ⬚C/s samples, to capture bainite

1628—VOLUME 33A, JUNE 2002 METALLURGICAL AND MATERIALS TRANSACTIONS A

 (a) (b) (c)
Fig. 2—Schematic of procedure for simulating Mn segregation. (a) Sheets of 5140 steel with low (0.82 wt pct) and high (1.83 wt pct) Mn levels are alternately
stacked into compacts. (b) Compacts are sealed in stainless steel boxes. (c) Boxes are rolled at 1150 ⬚C to form solid plate with compositional gradients.

Fig. 4—Time and temperature schematic of interrupted cooling techniques

used to observe microstructural development. Samples were austenitized
at 850 ⬚C, cooled at 1 ⬚C/s and 0.1 ⬚C/s, and quenched from progressively
lower temperatures to capture ferrite and pearlite growth in situ.

and high-Mn bands. Note that the high-Mn bands are always
on the right-hand side of the micrographs. In the sample
cooled to 750 ⬚C and quenched (Figure 5(a)), the microstruc-
ture consists of martensite in both bands. A slight difference
in etching response is the only distinguishable feature
between the different alloys.
The temperature at which austenite begins to transform
Fig. 3—Photographs of processing stages for simulating Mn segregation. to ferrite, i.e., the Ar3 temperature, is approximately 730 ⬚C
(a) Compact of alternately stacked low (0.82 wt pct) and high (1.83 wt for the low-Mn alloy at this cooling rate. This temperature
pct) Mn sheets and (b) stainless steel box.
was determined from inspection of specimens quenched
from between 750 ⬚C and 680 ⬚C.[22] Figure 5(b) shows large
growth. The temperature of the specimens was monitored amounts of ferrite in the low-Mn band at 680 ⬚C. Carbon
with a type-K thermocouple. Samples were prepared for is rejected from the ferrite allotriomorphs and, with a further
light optical microscopy with standard grinding and pol- decrease in temperature to 660 ⬚C (Figure 5(c)), pearlite
ishing procedures. Etching was performed with a combina- begins to form in the low-Mn band. Figure 5(d) shows
tion of 2 pct nital and 4 pct picral. austenite decomposition to be almost complete in the low-
Mn band at 650 ⬚C. Also noticeable in this micrograph is
III. RESULTS the growth of an almost continuous pearlite band at the
interface. Thus, carbon has been enriched not only in the
A. Specimens Cooled at 1 ⬚C/s pearlite of the low-Mn band but also locally in the high-Mn
Figure 5 shows light optical micrographs of samples with band. Figure 5(e) shows that at 620 ⬚C, the interfacial band
160 ␮m bands cooled at 1 ⬚C/s and quenched from progres- of pearlite has grown to a well-defined layer. At 550 ⬚C
sively lower temperatures. The micrographs show micro- (Figure 5(f)), the interfacial pearlite has grown approxi-
structural evolution on cooling at interfaces between low- mately 20 ␮m into the high-Mn band. Figure 5(g) shows

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 33A, JUNE 2002—1629

Fig. 5—Microstructural development as a function of temperature for 5140 steel with 160-␮m segregated bands cooled at 1 ⬚C/s and quenched from (a)
750 ⬚C, (b) 680 ⬚C, (c) 660 ⬚C, (d ) 650 ⬚C, (e) 620 ⬚C, ( f ) 550 ⬚C, (g) 350 ⬚C, and (h) 25 ⬚C. Light micrographs, nital and picral etch.

that bainite, along with martensite which has formed from ferrite and pearlite in the low-Mn band, martensite and bain-
austenite that was untransformed at 350 ⬚C, is present in the ite in the high-Mn band, and a 20-␮m-thick band of pearlite
high-Mn band at 350 ⬚C. The sample continuously cooled in the high-Mn bands adjacent to the interface between the
to room temperature (Figure 5(h)) consists of proeutectoid high- and low-Mn bands.

1630—VOLUME 33A, JUNE 2002 METALLURGICAL AND MATERIALS TRANSACTIONS A

Fig. 5—Continued. Microstructural development as a function of temperature for 5140 steel with 160-␮m segregated bands cooled at 1 ⬚C/s and quenched
from (a) 750 ⬚C, (b) 680 ⬚C, (c) 660 ⬚C, (d ) 650 ⬚C, (e) 620 ⬚C, ( f ) 550 ⬚C, (g) 350 ⬚C, and (h) 25 ⬚C. Light micrographs, nital and picral etch.

Figure 6 shows the microstructural evolution in samples cementite forms as both spherical particles and lamellae in
with 20 ␮m bands cooled at 1 ⬚C/s and quenched at progres- the low-Mn band. The discrete cementite particles appear
sively lower temperatures. The sample cooled to 750 ⬚C and to be related to a divorced eutectoid reaction, and their
quenched (Figure 6(a)) consists of martensite in both bands. formation is consistent with the growth of spherical, i.e.,
The martensite in the high-Mn bands etches darker than that divorced, cementite particles and ferrite directly from austen-
in the low-Mn bands. Figure 6(b) shows that proeutectoid ite at low undercoolings.[23]
ferrite is present in low-Mn bands of the sample cooled at In the sample cooled to 660 ⬚C (Figure 7(f)), the transfor-
690 ⬚C. Larger fractions of proeutectoid ferrite can be seen mation to proeutectoid ferrite and pearlite is complete in the
in the samples cooled to 660 ⬚C and 650 ⬚C (Figures 6(c) low-Mn band. The interfacial band of pearlite has grown to
and (d)), respectively. Note that some of the ferrite grains a thickness of approximately 20 ␮m. Figure 7(g) shows the
have grown to 20 ␮m in size, thus fully converting sections interfacial pearlite layer to be approximately 30-␮m thick
of the low-Mn bands to low-carbon-content ferrite. Pearlite at 650 ⬚C. Pearlite colonies separate from the interfacial
is also present in these samples and appears to be growing layer form at 640 ⬚C in the center of the high-Mn band, as
into the high-Mn bands. At 630 ⬚C (Figure 6(e)), pearlite indicated at by point A in Figure 7(h). The discrete pearlite
has grown further into the high Mn-bands, and at 600 ⬚C colonies, together with small amounts of proeutectoid ferrite,
(Figure 6(f)), the transformation of austenite is complete. are more prominent in the high-Mn bands at 630 ⬚C (Figure
Low-Mn bands are proeutectoid ferrite and pearlite and high- 7(i)), and the transformation appears to be complete at
Mn bands are pearlite. Note the striking contrast between 620 ⬚C (Figure 7(j)). The final microstructure consists of
the final microstructures of the 160 and 20 ␮m samples approximately equal amounts of proeutectoid ferrite and
cooled at this rate. Samples with 160 ␮m bands have mar- pearlite in the low-Mn band and pearlite with a small fraction
tensite and bainite in high-Mn layers; the 20 ␮m samples, of proeutectoid ferrite in the high-Mn band. The layer of
in comparison, contain only pearlite in high-Mn layers. interfacial pearlite is approximately 30-␮m thick.
Figure 8 shows samples with 20 ␮m bands cooled at
0.1 ⬚C/s. In the sample cooled to 750 ⬚C and quenched
B. Specimens Cooled at 0.1 ⬚C/s (Figure 8(a)), the microstructure of both bands is martensite.
Figure 7 shows light micrographs of samples with 160 Proeutectoid ferrite forms in the low-Mn band at 720 ⬚C
␮m bands cooled at 0.1 ⬚C/s and quenched from progres- (Figure 8(b)), and austenite continues to transform to ferrite
sively lower temperatures. The high-Mn bands are always at 700 ⬚C (Figure 8(c)). Pearlite forms in the low-Mn band
on the right-hand side of the micrographs. The sample cooled at 685 ⬚C (Figure 8(d)), but is limited to formation only in
to 750 ⬚C (Figure 7(a)) has a microstructure of martensite the carbon-enriched austenite that has not transformed to
in both bands. The Ar3 temperature was estimated to be ferrite. At 670 ⬚C (Figure 8(e)), pearlite begins to grow
740 ⬚C for low-Mn SAE 5140 at this cooling rate.[22] Figure into the high-Mn bands, and at 660 ⬚C (Figure 8(f)), the
7(b) shows that a large fraction of proeutectoid ferrite has transformation is complete. Low-Mn bands are almost
formed in the low-Mn band at 710 ⬚C. Pearlite colony nucle- entirely ferrite and high-Mn bands are pearlite. Once again,
ation and growth has commenced at 690 ⬚C (Figure 7(c)), note the striking difference in the final microstructure of
and at 685 ⬚C (Figure 7(d)), pearlite colonies appear at the samples cooled at this rate. Figure 7(j) shows that low-
interface between segregated bands. Figure 7(e) shows that Mn bands in 160 ␮m samples contain approximately equal
the interfacial pearlite is more prominent at 670 ⬚C. Colonies amounts of ferrite and pearlite. In contrast, Figure 8(f) shows
have merged into a discrete band of pearlite approximately low-Mn bands to be almost entirely ferrite in the 20 ␮m
5-␮m thick. It also appears at this cooling rate that the samples.

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 33A, JUNE 2002—1631

Fig. 6—Microstructural development as a function of temperature for 5140 steel with 20-␮m segregated bands cooled at 1 ⬚C/s and quenched from (a)
750 ⬚C, (b) 690 ⬚C, (c) 660 ⬚C, (d ) 650 ⬚C, (e) 630 ⬚C, and ( f ) 600 ⬚C. Light micrographs, nital and picral etch.

IV. DISCUSSION steel with Mn segregation. Complex layered microstructures

formed, and the capacity to form martensite upon cooling,
The micrographs in the previous section show that the i.e., hardenability,[24] decreased in specimens with fine
segregated band thickness has a striking effect on the contin- bands, as demonstrated by the samples cooled at 1 ⬚C/s.
uous cooling transformation (CCT) behavior of low-alloy Figure 5(h) shows that samples with 160 ␮m bands contain

1632—VOLUME 33A, JUNE 2002 METALLURGICAL AND MATERIALS TRANSACTIONS A

Fig. 7—Microstructural development as a function of temperature for 5140 steel with 160-␮m segregated bands cooled at 0.1 ⬚C/s and quenched from (a)
750 ⬚C, (b) 710 ⬚C, (c) 690 ⬚C, (d ) 685 ⬚C, (e) 670 ⬚C, ( f ) 660 ⬚C, (g) 650 ⬚C, (h) 640 ⬚C, (i) 630 ⬚C, and ( j ) 620 ⬚C. Light micrographs, nital and picral etch.

a large fraction of martensite and bainite in the final micro- also affected. Samples with 160 ␮m bands cooled at 0.1
structure. In contrast, Figure 6(f) shows 20 ␮m samples ⬚C/s required approximately 3750 seconds to complete trans-
cooled at the same rate to be entirely ferrite and pearlite. formation. In contrast, samples with 20 ␮m bands required
Thus, a reduction in segregated band thickness reduced the only 2850 seconds.
capacity to form martensite, i.e., a reduction in band thick- As mentioned previously, the underlying cause of micro-
ness decreased hardenability. The transformation time was structural banding is compositional segregation in the form

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 33A, JUNE 2002—1633

Fig. 7—Continued. Microstructural development as a function of temperature for 5140 steel with 160-␮m segregated bands cooled at 0.1 ⬚C/s and quenched
from (a) 750 ⬚C, (b) 710 ⬚C, (c) 690 ⬚C, (d ) 685 ⬚C, (e) 670 ⬚C, ( f ) 660 ⬚C, (g) 650 ⬚C, (h) 640 ⬚C, (i) 630 ⬚C, and ( j ) 620 ⬚C. Light micrographs, nital
and picral etch.

of alternating high- and low-solute layers. However, as dem- specimens illustrates the effect of segregation on continuous-
onstrated by several studies, microstructural banding does cooling behavior. Figure 6(f) shows that the high-Mn bands
not always appear in steel with compositional gradi- in specimens with 20 ␮m spacings transform to 100 pct
ents.[3,20,25] For example, at cooling rates high enough to pearlite after being cooled to 600 ⬚C at a rate of 1 ⬚C/s. In
form martensite in both high- and low-solute bands, the contrast, the CCT diagram shown in Figure 9(b), assuming
underlying pattern of segregation can be hidden behind a uniform austenite chemistry, shows that austenite should be
uniform martensitic structure, i.e., high cooling rates sup- untransformed under these conditions (point A in the
press microstructural banding. At cooling rates slower than diagram).
the critical velocity to form martensite in both bands, micro- This dramatic effect of Mn segregation on CCT behavior
structural banding appears to be due to the effects of alloy is associated with the growth of a pearlite layer at the inter-
chemistry on the nucleation and growth of ferrite and face of low- and high-Mn bands. The formation of interfacial
pearlite. pearlite is due to the redistribution of carbon between high-
Figure 9 summarizes, as CCT diagrams, the microstructures and low-Mn bands.[18,19] Ferrite allotriomorphs in low-Mn
that form as a result of continuous cooling in the low- and bands reject carbon into the austenite of adjacent high-Mn
high-Mn bands. The CCT diagrams were determined from bands. Thus, the austenite of the high-Mn bands near the
specimens with coarse band spacings, where the transforma- interface is enriched in carbon to the level (at least 0.56 wt
tions progressed independently of interfacial effects.[22] There- pct, as shown in Figure 1) required for pearlite nucleation.
fore, the diagrams represent CCTs in steels of uniform The interfacial pearlite colonies form a continuous layer on
composition and do not show the effects of band spacing and the low-Mn bands that have completely transformed to ferrite
interface-chemistry changes on microstructural evolution. and pearlite. Further growth of the interfacial pearlite is,
A comparison of the CCT diagrams in Figure 9 to the therefore, dependent only on the carbon content of the high-
microstructures that developed in the artificially banded Mn band as temperature decreases. Figure 10 shows that

1634—VOLUME 33A, JUNE 2002 METALLURGICAL AND MATERIALS TRANSACTIONS A

Fig. 8—Microstructural development as a function of temperature for 5140 steel with 20-␮m segregated bands cooled at 0.1 ⬚C/s and quenched from (a)
750 ⬚C, (b) 720 ⬚C, (c) 700 ⬚C, (d ) 685 ⬚C, (e) 670 ⬚C, and ( f ) 660 ⬚C. Light micrographs, nital and picral etch.

100 pct pearlite can form from noneutectoid carbon concen- pearlite growth rates may also accompany the formation of
trations if austenite is undercooled to the hatched region of the pearlite layer in the 0.4 wt pct C austenite of the high-
the phase diagram, i.e., the region where austenite is unstable carbon bands with decreasing temperature,[28,29] but such
with respect to both ferrite and cementite.[26,27] Changes observations were outside the scope of this investigation.
in interlamellar spacing, cementite lamellae thickness, and The growth of the interfacial pearlite layer is limited by

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 33A, JUNE 2002—1635

 (a)
Fig. 10—Eutectoid region of the equilibrium phase diagram for 51xx steel
with 1.83 wt pct Mn showing austenite with 0.40 wt pct C can transform
to 100 pct pearlite if undercooled to the hatched area of the diagram,
i.e., the region where austenite is unstable with respect to both ferrite
and cementite.[21,26,27].

an interfacial pearlite layer. Consequently, the spacing of

the bands results in profound differences in continuously
cooled microstructures as a function of band thickness.
2. The growth of the interfacial pearlite layer is limited by
carbon diffusion, and the layer grows to a thickness of
only 20 or 30 ␮m, depending on cooling rate.
3. In specimens with coarse band spacings, the limited
growth of the interfacial pearlite layer leaves a consider-
able volume fraction of high-Mn austenite, which trans-
forms to microstructures of bainite and martensite.
4. In specimens with fine band spacings, the interfacial
pearlite layer grows to the full thickness of the high-Mn
bands, and no bainite or martensite forms.
(b)
Fig. 9—Estimated CCT diagrams for 5140 steel with (a) 0.82 wt pct Mn
and (b) 1.83 wt pct Mn showing the effect of Mn on transformation behavior. ACKNOWLEDGMENTS
P is pearlite, B is bainite, and M is martensite.
The research for this article was conducted at the Colorado
School of Mines as part of the requirements for a Masters
of Science thesis by one of the authors. Support for the
decreasing carbon diffusion with decreasing temperature work was provided by the Advanced Steel Processing and
and, at best, grows to 20 ␮m from the interface in specimens Products Research Center, a university-industry cooperative
cooled at 1 ⬚C/s and to 30 ␮m in specimens cooled at research center at the Colorado School of Mines. The labora-
0.1 ⬚C/s. Figures 5 and 6 show that austenite outside of the tory heats of steel were generously provided by the Timken
pearlite layer in the high-Mn bands transforms to bainite Company (Canton, OH). American Axle and Manufacturing
and martensite in the 160 ␮m samples, while in the 20 (Detroit, MI) provided support during the writing of the
␮m bands, the austenite in the high-Mn bands transforms manuscript.
completely to pearlite. Thus, alloying elements that are
added to enhance hardenability can also lead to inconsistent
microstructures due to segregation patterns created during REFERENCES
solidification and rolling. 1. W.F. Smith: Structure and Properties of Engineering Alloys, McGraw-
Hill, New York, NY, 1993, pp. 125-28.
2. R.E. Reed-Hill and R. Abbashian: Physical Metallurgy Principles,
V. CONCLUSIONS PWS-Kent, Boston, MA, 1992, pp. 459-61.
3. T.B. Smith, J.S. Thomas, and R. Goodall: J. Iron Steel Inst., 1963,
July, pp. 602-09.
1. At slow cooling rates, the formation of ferrite in low- 4. J.D. Lavender and F.W. Jones: J. Iron Steel Inst., 1949, vol. 163, pp.
Mn bands results in carbon enrichment in the austenite 14-17.
of adjacent high-Mn bands and causes the nucleation of 5. R.G. Ward: J. Iron Steel Inst., 1965, Sept., pp. 930-33.

1636—VOLUME 33A, JUNE 2002 METALLURGICAL AND MATERIALS TRANSACTIONS A

 6. R.M. Fisher, G.R. Speich, L.J. Cudy, and H. Hu: Physical Chemistry 19. J.S. Kirkaldy, J. von Destinon-Forstmann, and R.J. Brigham: Can.
in Metallurgy: Proc. Darken Conf., United States Steel, Monroeville, Met. Q., 1962, vol. 1, pp. 59-81.
PA, 1976, pp. 466-73. 20. L.E. Samuels: Optical Microscopy of Carbon Steels, ASM INTERNA-
7. E.T. Turkdogan and R.A. Grange: J. Iron Steel Inst., 1970, vol. 208, TIONAL, Metals Park, OH, 1980, pp. 127-28.
pp. 482-94. 21. B. Sundman, B. Jansson, and J.O. Anderson: CALPHAD (Calculation
8. R.D. Doherty and D.A. Melford: J. Iron Steel Inst., 1966, Nov., pp. of Phase Diagrams): A Comprehensive Guide, Elsevier Science, New
1131-43. York, NY, 1985, pp. 153-190.
9. T.Z. Kattamis and M.C. Flemings: Trans. TMS-AIME, 1965, vol. 233, 22. T.F. Majka: Master’s Thesis, Colorado School of Mines, Golden,
pp. 992-99. CO, 2000.
10. F. Weinberg and R.K. Buhr: J. Iron Steel Inst., 1969, Aug., pp. 1114-20. 23. J.D. Verhoeven and E.D. Gibson: Metall. Mater. Trans. A. 1998, vol.
11. R.A. Grange: Metall. Trans., 1971, vol. 2, pp. 417-26. 29A, pp. 1181-89.
12. G. Mukherjee: Tool Alloy Steel, 1977, vol. 11, pp. 55-58. 24. G. Krauss: Steels: Heat Treatment and Processing Principles, ASM
13. R. Grossterlinden, R. Kawalla, U. Lotter, and H. Pircher: Steel Res., INTERNATIONAL, Materials Park, OH, 1990, p. 152.
1992, vol. 63, pp. 331-36. 25. T.F. Majka, D.K. Matlock, G. Krauss, and M.T. Lusk: 42nd MWSP
14. S.W. Thompson and P.R. Howell: Mater. Sci. Technol., 1992, vol. 8, Conf. Proc., ISS, Warrendale, PA, 2000, pp. 75-87.
pp. 777-84. 26. R.F. Mehl and W.C. Hagel: Progr. Met. Phys., Pergamon Press, New
15. J.A. Eckert, P.R. Howell, and S.W. Thompson: J. Mater. Sci., 1993, York, NY, 1956, pp. 74-134.
vol. 28, pp. 4412-20. 27. D.A. Porter and K.E. Easterling: Phase Transformations in Metals
16. P.G. Bastien: J. Iron Steel Inst., 1957, vol. 174, pp. 281-91. and Alloys, Van Nostrand Reinhold, Berkshire, United Kingdom, 1981,
17. W.C. Leslie: The Physical Metallurgy of Steels, Tech Books, New pp. 333-34.
York, NY, 1981, pp. 172-73. 28. M.P. Puls and J.S. Kirkaldy: Metall. Mater. Trans, 1972, vol. 3, pp.
18. C.B. Voldrich: Suppl. J. Am. Welding Soc., 1947, vol. 26, pp. 2777-96.
153s-69s. 29. B.E. Sundquist: Acta Metall., 1968, vol. 16, p. 1413.

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 33A, JUNE 2002—1637

View publication stats