Lawrence Berkeley National Laboratory

Lawrence Berkeley National Laboratory

Title
HEAT TREATMENT AND CHARACTERIZATION OF DUPLEX 1010 AND
1020 STEELS

Permalink
https://escholarship.org/uc/item/3r81b4dn

Author
Young, M.J.

Publication Date
1977-11-01

eScholarship.org Powered by the California Digital Library

University of California
 _ l i i
- LBL-6620

TABLE OF CONTENTS

ABSTRACT . v

I. INTRODUCTION 1

II. EXPERIMENTAL PROCEDURE 5

A. Materials 'Preparation . 5

B. Heat Treatment 5

1. Intermediate Quenching 5

2. Continuous Annealing. . . . . . . . . . . . 6

C. Mechanical Testing. . 6

1. Tensile . . 6

2. Impact. 7

D. Dilatometry 7

E. Metallography 8

1. Optical 8

2. Transmission Electron Microscopy 8

3. Fractography. . . 9

F. Volume Fraction Determinations. . . . . . . . . . . . . . 9

III. RESULTS AND DISCUSSION 10

A. Microstructures'. •••• . . . . . . . . . . . . . . . . . 11

1. Intermediate Quenching vs. Continuous Annealing (lOlO) 11

2. Intermediately Quenched 1010 and 1020 14

- a. Martensite Morphology 14
.'. — NWitt ;
Ifcb riport * M p n p t t t d ^ i a iccounl of wort
b. Additional Microstructural Features 15
Unittd Sine* not tfie United Stuei DepMtment of
Ewriyt not toy or Atbr cmptoytet,nor w y of their
coniittioti, mheontnciore. or tnifr c m p f c j a a , o n t o
iny wuunty, u p n n . o r rinpiiid,ari*mnct iaf fcpl c. Tempered M i c r o s t r u c t u r e s 16
tiihllttyoT iwpomfcffity for thaicaiacy canipb(tnelt
t

n taifutnm of anytnfonmifaa,ippnitul, ptodiitt 6t

praamdi*doMd oi*rtptMtnb ihil In u * woutd not
t

infitan privately owned ttthti.

rjETOBUTlOfil QEIHIS QQCUMOfl. IS UNUSjgEQ

 rivS

d. Impact Fracture Surfaces 17

B. Mechanical Properties. 17

1. Tensile. 17
a. Intermediate Quenching vs. Continuous Annealing
(ioioy ...... . . . . • >. • .• .,.• • . . . . . . . J.*

b. Intermediately Quenched 1010 and 1020 18

1. As quenched 19

ii. Tempered 20

2. Impact . . . . >... • . ... ....-• • • . ....••;-... *'

a. tn^rmediately Quenched 1010 And 1020. 22

h. DBTT;(1020). 23

C. Mechanical Property/Micrqstructure Correlations, 24

1. Quantitative (Empirical) Relationships 24

2. Qualitative Relationships 29

p. Comparison To Commercial Specifications 31

3 3
IV. CONCLUSIONS • *
3 5
ACKNOWLEDGMENTS

REFERENCES . . . . . . . . . 36
3 9
TABLES • .->•••:••• • • •

FIGURE CAPTIONS ...'.;....,.. . , * • • . . . . . 43

FIGURES,f« 47
 -V-

HEAT TREATMENT AND CHARACTERIZATION OF DUPLEX 1010 AND 1020 STEELS

M. J. Young

Materials and Molecular Research Division

Lawrence Berkeley Laboratory
and Department of Materials Science and Engineering
University of California, Berkeley, California 94720

ABSTRACT

An investigation concerning the heat treatment of duplex

ferritic-martensitic 1010 and 1020 steels has been carried out.

Tensile and impact properties of intermediately quenched and contin­

uously annealed specimens have been characterized and microstructural

correlations established. The rule for two phase mixtures may be

used as a fairly good approximation of the mechanical behavior of

the duplex system, and improved correlations are possible with suitable

modifications for the size, shape and distribution of the second phase.

The property differentials between the duplex 1010 and 1020

at equal volume fractions of martensite are due mainly to the higher

carbon content (hence, higher strength) of the 1020 martensite.

Important metallurgical variables include subgrain formation, retained

austenite, transformation twinning and martensite fraction. Of these,

the volume fraction of martensite appears to be the most influential

in controlling the strength, ductility and impact properties of the

duplex systems studied.

 -VI

ACKNOWLEDGMENTS

I would like to thank Professor Gareth Thomas for suggesting

this research program, and for continuous guidance and encouragement.

I would also like to thank my many friends, particularly Lorin and

Mary Brass, who made the time I spent in Berkeley enjoyable. I

would especially like to express my appreciation to my good friend

Jayoung Koo, for his invaluable assistance in all phases of this

work.

This work was done under the auspices of the U. S. Department

of Energy through the Materials and Molecular Research Division of

the Lawrence Berkeley Laboratory.

 1

I. INTRODUCTION

The ever increasing emphasis on energy and resource conserva­

tion has triggered intensive development programs resulting in the

introduction of new steel compositions and processing techniques.

Many programs, e.g., weight savings in the transportation industry,

have concentrated efforts in the area of high strength, low alloy

(HSLA) steels which derive their principal strengthening from finely

dispersed Niobium, Vanadium or Titanium carbides. Many design goals

have been met with the HSLA steels, although formability appears to

be a problem and in numerous cases the alloying and processing

economics more than offset cost reductions due to the strength-to-

weight increase. Alternative emphasis, both commercially and

2-5
academically, toward duplex, low carbon steels has thus increased,

and results have been very promising.

The processing of steels to obtain the microduplex structure

of martensite in a dislocated ferrite matrix is not an entirely new

concept. The duplex ferritic-martensite aggregates can be obtained

in several ways, thus mechanical, thermal-mechanical and thermal

methods have been investigated. Cairns and Charles instituted an

early program aimed at developing laminates and composites through

combinations of mechanical work and heat treatment. They achieved

only limited success, especially with commercial steels, but did

propose a useful model for the fiber strengthening of steel. A viable

alternative utilizes annealing in the two phase field; however the

 2

microduplex structures cannot be formed in every combination of alloy

compositions and annealing temperatures. Diffusion characteristics

and such phenomena as reverse martensitic transformations prevent

that formation in some systems. Duplex processing concepts have been

applied to nonferrous as well as ferrous systems, but with considerably

more emphasis on the latter. Hwang and Jin et. al., have utilized

the (a + Y ) region in maraglng steels to obtain desirable properties

in cryogenic applications, and Enomoto and Furubayashi have studied

the vast two phase region In iron-nickel alloys. The beneficial effects

of refined grain size have been asserted by many researchers, and are

well understood -ud documented. Miller has shown that stable, ultra-

fine grained microstructures can be obtained by annealing cold worked

alloys at relatively low temperatures In the (« + Y) range. Similarly,

work by Snape and Church with low alloy steels has shown that the

presence of the newly formed phase at the grain boundaries of the

primary phase can effectively retard grain growth. The combined

strength increases from grain refinement and the presence of martensite

(which acts as the load carrying constituent), coupled with the

ductility inherent in the ferrite matrix has been utilized recently

by Koo and Koo and Thomas, using commercial 1010 and experimental
. , - ' ' ; • • • ' • • • • " ' • ' • '

- - 12
alloys. In their work a thermal process originated by Grange was
modified to include two phase annealing which significantly Increased

strength while maintaining good elongation ductility. Industrial

**"" 13
research with these duplex alloys has shown that significant
strengthening and work hardening can be obtained with thermal-mechanical
 3

treatment. After tempering, the martenslte-ferrite aggregate is of

intermediate strength with ductility equivalent to HSLA steels of

comparable strength.

Thus, the advantages of duplex microstructures have been

established in several alloy systems. Applications to the low carbon

steels are especially interesting for several reasons: (1) composi­

tions and required heat treatments are basically simple, with the

possibility of eliminating the need for thermal-mechanical processing,

(2) a very high tonnages are used commercially and (3) considerable

scope for increasing strength while maintaining good ductility appears

possible. This has been a driving force for the ongoing duplex steel

design program in Professor Thomas' group. It is aimed at char­

acterizing the duplex mlcrostructure/mechanical property relations

with particular emphasis on further improvements in mechanical

properties, and a fundamental understanding of their origins. As part

of this design program, the present research was intended to establish

structure-property relationships in high purity 1010 and 1020 steels

(0.1 and 0.2 weight percent carbon, respectively) and precedes

selected ternary and quaternary additions to the base alloy. The

immediate research objectives were:

(1) Determine the effects of carbon, in the separate

phases as well as in the overall composition.

(2) Characterize the duplex microstructure in terms

of martensite and ferrite quantities, and establish

the mlcrostructure-property relations.

 (3) ^Determine if further mechanical property^improve­

ments, solely be heat treatment, would be possible.

Results of the previous work by Koo ' have been utilized;

sp^eific.^y .,tbAt ,,Qyc.ling:.ia.-the/i(a '+ y) region more than" once does

lc :
T

not result, in further grain refinement nor substantial improvement

in mechanical properties. Investigations concerning processing to

obtain the duplex microstructure, tempering effects and impact

behavior also have been completed.

f
* '
<- 1 V rij

j
f ll' ' MS
a
=-j
1
, ->LI 3 !
!
><'r>
,^f
\ t * ^ r
'?A '••> *<r I 1
 5

II. EXPERIMENTAL PROCEDURE

A. MATERIALS PREPARATION

High purity 1010 and 1020 alloys of the composition shown in

Table 1 were vacuum arc remelted and cast as 20 pound ingots. These

ingots were subsequently upset forged to approximately 0.5 in. thick­

ness. Portions were further reduced to approximately 0.15 in. by

cold rolling.

B. HEAT TREATMENT

1. Intermediate Quenching.

Oversize flat tensile and Charpy impact specimens were

austenitized in a vertical tube furnace (resistance heating) under

an argon atmosphere for 30 minutes at 1100 °C, and quenched directly

into agitated iced brine. Microhardness measurements were taken

across numerous cross sections and compared to values obtained by

other researchers. ' ' Later transmission electron microscopy

showed the structures to be almost fully martensitic and free of

undissolved carbides. Further thermal treatments were accomplished

according to the following schedule:

a. Two Phase Annealing.

Oversize tensile and Charpy specimens were reheated to the

desired temperature In the (a -I- Y ) field and held for 20 minutes,

the vertical furnace with protective argon atmosphere was again used.

The specimens were then quenched directly into iced brine. Extensive
 6

:!;
optical metallx)gra^hy'-^eye'id:e&^ Mimost no evidence of texture resulting

from the forging or cold rolling operations,

b. Tempering^

. Tempering.of the martensitlc and duplex alloys was effected

by immersing the samples in a neutral salt bath at 200 °C for 30

minutes. Selected samples were also tempered at 400 °C and 600 °C

for 30.minutes.

2. Continuous Annealing.

Oversize flat tensile specimens of the 1010 alloy were

encapsulated in quartz tubes which were then evacuated and backfilled

with 20 inches of argon. These were austenitized in a box furnace

at 1100 °C for 30 minutes and then transferred directly to an adjacent

furnace at the desired two phase temperature and held for 20 minutes.

The specimens were then directly quenched into agitated iced brine;

the quartz capsules were imploded simultaneously with immersion in

the iced brine to ensure rapid quenching.

C. MECHANICAL TESTING

1. Tensile.

Tensile properties were determined using the flat tensile

specimens shown in Figure 2A. The specimens were originally machined

oversize, heat treated and ground to final dimensions. Approximately

0.025 in. was removed from each side to eliminate the possible effects

of surface decarburization. Optical metallography revelaed

decarburization did not exceed approximately 10 mils.

 7

-Tensile tests were performed at room temperature on a

calibrated Instron machine using a cross head speed of 0.05 cm/min

and full scale load of 1000 Kg. Uniform elongation was continuously

monitored with an optical magnifier accurate to 0.001 in. Total

elongation was determined by measuring gage lengths before and after

testing with an optical microscope equipped with a vernier translating

stage accurate to 0.001 in.

2. Impact.

Impact properties were determined using the standard and

subsize Charpy V-notch specimens shown in Figure 2B. The low hard-

eijiablity of the 1010 alloy precluded the use of fullsize specimens,

and subsize 1020 specimens were included for comparison. Impact

•tests were performed using a Universal Impact machine with calibrated

units of 0.25 ft. lbs. Low and high temperature testing was,conducted

according to ASTM specifications. Zero and sub-zero temperatures

were controlled using mixtures of ethyl alcohol and dry ice.

D. DILATOMETRY

Dilatometric methods were used to determine the austenite

start, A , and austenite finish, A,, temperatures as a prelude to

determining the A, and A. temperatures for the 1010 and 1020 steels.

Martens!tic transformation temperatures, M and M , for the completely

f

martensitic and duplex alloys also were determined.

 8

E. METALLOGRAPHY

1. Optical.

Sections to be used for optical metallography were cut from

fractured tensile and Impact specimens and mounted in Koldmount,

Bakelite or Lucite. The specimens were mechanically ground to 600

grit, ultrasonically cleaned and polished on a 1 micron rotating

diamond abrasive wheel lubricated with kerosene. The standard etchants

2% Nital, 5% Nital and Railings reagent were used to reveal the

ferritic and martensitic structures. Prior austenite grain boundaries

were revealed by immersion in"an etch solution of 1 g picric acid,

1 g dodecylbenzene' sulfonate and 100 ml of water. Samples were

examined on the Zeiss Ultraphot il metallograph equipped with a

Nomarsky interference contrast apparatus.

2. Transmission Electron Microscopy.

,' Specimens of approximately 0.150 in. thickness for thin foils

. were treated with the tensile and impact specimens, and carefully

ground to 30 mils. Approximately equal amounts were removed from

each side while utilizing flood cooling to minimize specimen heating.

The specimens were then chemically thinned to less than 10 mils in

' a solution of H»0_ containing 2% HP at room temperature. 3.0 mm

"-/ dto'cB'were spark cut and'mechanically sanded to 1-2 mils, cleaned

with acetone and electropolished in a twin jet polishing apparatus

* at room temperature using a chromic-acetic solution '

 9

(75 g Cr0 + 400 ml CHjCOOH + 21 ml distilled HjO).

3 Polishing times

varied from 3 to 5 minutes at 30-35 ma and 25-30 volts. Foils were

examined in a JEM 7A electron microscope at an accelerating voltage

of 100 KV.

3. Fractography.

Fracture surfaces of tensile and impact specimens were examined

using an AMR. 1000 scanning electron microscope with secondary emission

at 25 KV. Specimens were cut, ultrasonically cleaned and stored in

dessicators until examination.

F. VOLUME FRACTION DETERMINATIONS

Two methods were used in the martensite fraction measurements:

18
the linear intercept method, and the systematic count method using

a scribed eyepiece. Measurements were made on many optical micographs

and fields of view, on Charpy and tensile fractures and at various

magnifications. Values using the two methods were averaged and

rounded to the nearest 5%. .

 10

III. RESULTS AND. DISCUSSION

Heat treatment in the two phase field does not necessarily

involve equilibrium conditions, and the temperatures, the line

(lever rule) and composition determinations as specified by the

equilibrium diagram are not necessarily applicable. The determina­

tion of the A., (eutectoid) and A 3 (y solidus) temperatures for the

exact .composition was thus necessitated, and extensive experimental

heat treatments near the A and A f temperatures (from dilatometry)

were conducted. Based on optical metallography, the A^ temperatures

were determined to be ^ 5 5 °C for 1010 and ~815 °C for 1020. The

A ,for both alloys} was ^710*0. The amount of phase separation

corresponding to 20 minutes at the holding temperature was determined

in the same manner.

Qualitatively, the lower limit of the acceptable range of

martenBite volume fractions (V ) is restricted to the value where no

m
strengthening of the duplex alloy occurs. The upper limit is set

when the failure of the martensite immediately leads to the failure

of the duplex structure, i.e., ferrite is not present in'.sufficient

quantity to contribute to toughness and ductility. Of course, the

final range of V will be determined from specifications and applica­
nt
tions for which the steel will be used. Important parameters may

.include strength, ductility, impact resistance (toughness), formability,

' weldability, etc.

•J >-.
 11

One of the more important (and difficult to arcumvent) results

of the non-equilibrium conditions concerns the composition of the

martensite in the duplex structure. The low atomic weight of carbon

prevented its chemical analysis by x-ray spectroscopic means, e.g.,

the EDAX attachment of the SEM. Diamond pyramid microhardness

measurements of the martensite in the continuously annealed material

appeared to show a valid correspondence (nonlinear) to the volume

fraction, which is inversely proportional to carbon content at a

given temperature. The fineness of the structure of the intermediately

quenched material prohibited valid testing: indentations resulted

in the deformation of the surrounding ferrite. The microhardness/volume

fraction relationship will be discussed in a later section. One

equilibrium relationship is that at equal volume fractions, i.e.,

different tie lines, the martensite in the 1020 will have a higher

carbon content than the martensite in the 1010 (see Figure 1). As

will be shown, experimental evidence supports the validity of this

relationship for the duplex system.

A. MICROSTRUCTURES (OPTICAL, TEM, FRACTOGRAPHY)

1.. Intermediate Quenching vs. Continuous Annealing (1010).

The-.initial microstructure is important in the duplex proc­

essing of steel, and martensite has been shown to be favorable

because of the numerous fine heterogeneities in the structure. *

Figure 3A shows the inartehsitlc structure of the steels before being

 12

subjected to the duplex treatments of Figure 1. In the intermediate

quenching process, reheating to the two phase temperature after

quenching from austenite results in the nucleation of austenite at

prior austenite grain boundaries, martensite lath boundaries, dis­

locations and other quenched-in imperfections. The high density of

aucleacion sites coupled with the slower reaction rate results in a

fine dispersion of martensite 'islands' in the ferrite matrix as

shown in Figure 3, B and D.

In the continuous annealing process (Figure 1C), the structure

prior to two phase annealing is austenite. Decreasing the temperature

to the two phase region results in several changes necessary for the

nucleation of ferrite: there is a fluctuation in carbon content and

:
i- 23 '
a local change in crystal structure. The ferrite nucleates at the

prior austenite grain boundaries because the surface energy barrier is

reduced, 23-25 and grows into the austenite matrix. The microstructure

thus depends upon prior austenite grain size and tends to be coarser

than that obtained from martensite as shown in Figure 3C.

The nucleation and kinetic differences in the two processes

were most evident in micrographs with increasing percentages of

martensite. In the intermediately quenched material, the existing

martenBlte islands did not grow appreciably, but more austenite

1
•*. ^-i" i \ \ f *

„ (martensite after quenching) nucleated. In the continuously annealed

"specimens, the number of nucleation sites did not appear to change,

.and the'martensite regions grow significantly.

lft
 13

Several striking differences in the microstructural aspects

of the two processes were evident in TEM examination. The ferrite

regions in all the specimens were locally heavily dislocated as a

result of accommodating the martensite transformation strain, but

the significant difference in proeutectoid morphologies was subgrain

size. The ferrite grains in the continuously annealed specimens were

very large, and there was no observable division into subgrains. The

proeutectoid ferrite in the intermediately quenched material had a

network of subgrains spaced an average of 0.5 to 1 micron apart.

The ferrite region in this process does not undergo a phase transforma­

tion as the alpha (martensite) is annealed in the two phase region.

As a result, the dislocations generated by the martensitic transforma­

tion (on quenching from 1100 °C) rearrange themselves into dislocation

walls and form subboundaries in the proeutectoid constituent during

the two phase annealing. Figures 4 and 5 illustrate the subgrain

formation, which is regarded as an important strengthening mechanism.

26
Warrington reported that subgrain size can be related to the flow

stress of iron by the Hall-Petch equation.

In the continuously annealed material almost all martensite/

ferrite interfaces were microscopically regular and smooth. While

smooth interfaces were evident in the intermediately, quenched specimens,

martensite laths (Figure 4) and regularly shaped regions (Figure SA)

were very prevalent. The ferrite in the immediate vicinity of

martensite/ferrite interfaces showed numerous regions of fibrous

carbide precipitation resembling pearlite as shown in Figure 6.

 14

As is evident from the figure, such carbidesi were present from both

processes. The carbides presumably remain after the moving ferrite/

austenite boundary grew into the matrix after the nucleatidn of the

Y •*• ot + carbide reaction. There appeared to be a higher and more

localized density of the fibrous carbides in the continuously annealed

specimens. Positive identification of the carbides was not possible

because of the complexity of the diffraction patterns from such areas.

The morphology of such carbides is similar to that observed in simple

2 7 •'•••'• " •

chromium steels by Campbell and Honeycombe who identified the car­

bides as H_C,, produced by pearlitl'c type of reaction.

Scanning electron fractographs of thetensile failures for

the two processes are shown in Figures 7 and" 8. Figure 7A shows the

the predominantly dimpled rupture surface of sample IB which is

indicative of a completely ductile failure. Figure 8 shows the

fracture surfaces of the continuously annealed materials at various

magnifications. Evidence of ductile failure is apparent, but areas

of cleavage failure are equally evident in A and B. High magnification

views of separate areas (C and D) emphasize the two'failure modes.

Reduction in area measurements were not made, but the continuously

annealed material failed after considerably less necking than the

Intermediately quenched specimens.

, 2. Intermediately Quenched 1010 And 1020.

a. Morphology of Martensite.

' ^The characteristic features of the martensite in the 1010 and

1
1020 steels after directly quenching from 1100 "C can be seen in
 15

Figures 9 and 10. The microstructure consisted almost entirely of

28
laths 0.1 to 0.3 microns wide. Krauss and Marder suggest that each

lath forms directly from an independent, homogeneous shear, and

successive shears produce a packet. The broad faces of the laths were

often visible, thereby allowing the packet boundaries to be delineated,

as in Figure 9A. The lath boundaries are frequently of low angle

29 30
character, whereas the packet boundaries are usually high angle. '

The morphology of martensite in duplex specimens did not differ

significantly from the fully raartensitic specimens. Figure 9C shows

the complex intersection of laths In specimen IB, and Figure 4 shows

a martensite region embedded in the ferrite in specimen 2B.

Another prominent feature of the martensite was the occurrence

of autotempering. The autotempered carbides were of the highest

density in the completely martensitlc structures but were evident in

the ferrite of duplex alloys also. The carbide diffraction spots

were very weak, so the carbides were tentatively identified by trace

analysis. Most of the carbides were cementlte In the typical {110}

Uidmanstatten orientation shown in Figure 10. The carbide morphology

Was not observed to vary with V .

m
b. Additional Microstruetural Features-

The observation of retained austenite was not unexpected since

its presence in such low carbon steels had been detected earlier by

Koo. ' The extremely thin films of remained austenite that were

trapped between growing martensite laths were Identified by indexing

selected area diffraction patterns such as the one given in Figure 11C.
 16

Figure' 11A is the bright field and Figure 11B is the dark field where
the' retained austenite reversed contrast when the (002) reflection
was -imaged. 'Figure llD shows the'indexed pattern which proves the
martensite and austenite exhibit the Kuriljumov-Sachs orientation
relationships, U l l H K l l O ) ' .
Very fine internal twins such as shown in Figure 12 were
observed in all of the duplex specimens, and are indicative of lo­
calized carbon* enrichment. While no intentional effort to establish
the amount b£ twinning was made, the twinning was more evident at
lower martensite fractions and in the 1020 foils.
The descriptions of ferrite and carbide morphologies for the
intermediately quenched 1010 apply also to the intermediately quenched
1020 steel.. As is evident from Figure 7, the fracture surfaces for
the two intermediately quenched alloys were also very similar.

C. TEMPERED MICROSTRUCTURES

Tempering at 200 °C is of considerable practical interest

because the steel sheet Is very likely to encounter such conditions
during processing, e.g., galvanizing, painting, etc. Tempering of
the 1010 and 1020 at 400 and 600 °C 1B of comparativelsr less practical
i
Importance and was carried out mainly to investigate the property
••-j i. trends. An effort to investigate the microstructural aspects after
> . it* * i
t tempering was thus not attempted. A detailed microstructural analysis
J 1 l T
i 32
'- of tempering of the duplex structure is forthcoming.
 17

d. Impact Fracture Surfaces.

The fracture surfaces of the duplex alloys were very In­

teresting because at most temperatures they showed localized dual modes

of failure. Ductile failure, as evident by dimpled rupture, and the

characteristic river patterns of cleavage are apparent in Figures 13

and 14; most failures could then be characterized by quasicleavage.

There was some isolated evidence of secondary cracking, as in Figure

14C. An analysis to determine unambiguously the origin of such areas

as in Figure 13A, e.g., by polishing and etching a plane normal to

32
the fracture surface, is in progress.

B. MECHANICAL PROPERTIES

1. Tensile.

a. Intermediate Quenching vs. Continuous Annealing (1010).

A comparison of mechanical properties for the two treatments

is aided by examination of Figure 15. A striking observation is that

the strengths of the continuously annealed material appear to be

very insensitive to changes in the martensite percentage. This may

be considered more unusual because the elongations are equally as

sensitive as these for intermediate quenching. Below approximately

40% martensite the strengths for continuous annealing are superior

to those for intermediate quenching. Examination of the optical

micrographs (Figure 3) assists in the explanation. Maintaining equal

volume frac'iions. at, low values of V the martensite. in the interme-

• i • . m

diately quenched materials is very uniformly and finely dispersed.

 18

The marfcensite in the continuously annealed specimens is uniformly

J

dispersed also, but the much larger regions are more interconnected

and thus carry more of the load than the intermediately quenched

martensite. This advantage can be seen to be decreasing as V ffl

increases, and past ~40% the strength of the intermediately quenched

material is greater, prasumably due to the finer martensite dispersion

and strengthening of fernte due to subgrain formation. As mentioned

previously, the elongation values for the continuously annealed

specimens are sensitive to V . Uniform elongations are 2% lower and

total elongations are 4% lower than for intermediate quenching, and

' the variance changes little over the range of V m tested,

b. Intermediately Quenched 1010 And 1020.

A comparison of the duplex 1010 and 1020 alloys is especially

' interesting- because the effects of carbon in the separate phases,

as well as in the total alloy content, may be very pronounced. The

prior austenite grain sizes of the duplex steels were compared to the

P " standard ASTM charts and determined to be ASTM #5 (7 to 8 microns)

for all duplex treatments. 'Therefore, this grain size, one of the

v*% * most Important factors in controlling the strength and associated

16
-<, J, ductility of low carbon steels, may be eliminated as a variable.

'• - " The 1010 and 1020 steels were subjected to the two phase

^ *-"' annealing processes shown" in Figure 1, A and B. Admittedly, the

" -^ >t ih6 id±nfe;tim bf 20 minutes in tne two'phase region does not result in
r 1
e

'•* i - » the-phase separation'dictated by the equilibrium diagram, but because

f
.>.'.&. ' the Holding time was constant, any deviations'from equilibrium are

"V, - ,, '
 19

expected to be relatively small. For a given alloy, the two phase

temperatures used to obtain differing volume fractions of martensite

were sufflcently close, e.g., 15-20 °C, that no marked kinetic

differences would be expected. One relationship evident from equilib­

rium, as discussed previously, Is that at equal volume fractions the

martenslte in the duplex 1020 steel has a higher carbon percentage

than the martensite in the 1010. Carbon content is the most important

factor in determining martensite strength, as has been documented by

. 14-16
many researchers.

i. As Quenched

When the duplex specimen is subjected to standard'tensile

testing, there are two stress regions of concern. From the beginning

of the test and until the yield stress is reached, the primary deforma­

tion occurs in the ferrite because it is the continuous phase. It

can then be seen that for a given V and morphology that the strength

of the martensite does not play an important role In determining the

yield strength. The carbon contents of the ferrite in the 1010 and

1020 are essentially equal, so there should be little difference In

the yield strengths of the two steels. As Figure 16 shows, the yield

strengths of the 1010 and 1020 differ by a maximum of ~5 ksi. Given

the limited accuracy of the volume fraction and strength determinations,

It can be seen that the yield strengths are very nearly equal, in

agreement with the previously mentioned considerations of continuum

mechanics.
 20

| M :
- The deformation" of the martensite is initiated shortly after
( 1
' "plastic flow begins and is accelerated at the onset of localized
deformation (necking)'!' The larger differences in the ultimate
' strengths of the* two alloys reflect the differences in martensite
strength due to carbon enrichment. The difference increases noticeably
as the volume fraction increases because there is less ferrite present
and the differences in the strengths of the carbon enriched martensites
are accentuated. As of course are the differences in work hardening.
The preceding arguments may be used to explain the elongation
results also. Uniform elongation is measured until necking occurs,
and mainly reflects the deformation of the ferrite. As can be seen
from Figure 16, the uniform elongations are nearly equal (0.5% maximum
differences). Total elongation encompasses the entire stress region
¥

of martensite deformation, and the larger differences between the

ldlO and'i020 are again reflective of the martensite strength (and
•' ductility) differences.
As expected, the common relationship of decreasing ductility
with Increasing strength is observed for both alloys.
l,s
"' ii:' Tempered
The tensile properties of the two steels after tempering at
J
- 200 °C are'shown as'functions of the fraction of martensite in Figure
17. ' The relative positions of the yield curves are reversed from
" the untempered condition; a situation not to be expected. This is
possibly an indication of the tolerances in tensile values and martens-
i

ite volume determinations. The small fluctutation in martensite

 21

percentage from sample to sample (as evidenced optically) is included

in the latter uncertainty.

The properties of the fully martensitic materials did not

change appreciably on tempering at 200 °C, mainly attributable to

the high density of autotempering (quench tempering). The ultimate

strengths of the fully martensitic structures increase slightly after

tempering, presumably due to disperison hardening offset (not quite

equally) by a depletion of carbon in solid solution. Small increases

may also be attributed to increased carbond segregation to, and

22
locking of, dislocations.

The duplex steels had significant strength decreases after

200 °C tempering, as shown in Figure 18. Carbide precipitation did

not occur in the ferrite during quenching presumably because there

was unsufficient time for carbon to diffuse over the large distances

involved. During the 30 minutes at 200 "C carbon had sufficient

time to diffuse into the depleted ferrite regions leading to carbide

precipitation.

The yield and ultimate strengths of the duplex alloys appear

to have an equal sensitivity to tempering at 200 °C, and strength/

elongation ductility relationships (Figure 23) appear to be changed

from the as quenched condition.

The strengths of the 1010 and 1020 steels containing 35 and

40% martensite, respectively, in the as quenched and tempered con­

ditions are shown in Figure 18. The alloys respond similarly to

tempering at 200, 400 and '600 °C. As expected, the ultimate strengths
 22

are more sensitive than the yield strengths. The elongation curves

know that the ductility (especially as evidenced by the total elonga­

tion) of the 1020 is more ''influenced than that of the 1010, and the

ductility in general is influenced more by tempering than is the

strength.

2. Impact.

a. Intermediately Quenched 1010 and 1020.

As previously mentioned, the low hardenability of the 1010

prohibited the use of full size charpy impact specimens. Preliminary

tests established that the maximum thickness that could be quenched

to a fully martensltic structure was approximately 190 mils, so the

subsize specimens shown in Figure 2 were used. Standard size specimens

were used for the 1020 steel, and subsize were also included for

comparison.

In low carbon steels, increasing the Fearlite content increases

33
the strain hardening rate and decreases the radius of an advancing

fibrouB crack21 in materials containing fractions of the hard second

r
phase. This increases the degree of stress intensification ahead of
r
•' ' " ' ' '-21'
advancing tear and facilitates cleavage. The situation in the

duplex structure may be analogous. The hard, second phase martensite

increases the stresB intensification and decreases the amount of

energy absorbed, as is clearly shown in Figure 19. The increase in

martensite volume fraction from 25 to 50% decreases the impact energy

(especially in the full-size 1020), presumably due to the aforementioned

reasons. The increasing cleavage as V. increases was also evident in

 23

the SEM study of the fractured specimens.

There is some variance in curve shape from 1010 and 1020, and

fullsize to subsize, but an essentially similar behavior is suggested.

Comparing the magnitudes of subsize full-size is incorrect, because

the failure mode is surely affected by specimen geometry. The dif­

ferent slopes of the 1020 subsize vs. fullsize curves may be indicative

of this. Concerning subsize specimens though, 1010 specimens showed

a quite pronounced shear lip at all martensite fractions while the

•1020 did only for the fully martensitic structure. The difference

in impact energy is evident from Figure 19, and an obvious conclusion

is that at equal volume fractions (but different strengths) 1010 is

more resistant to crack initiation (which the charpy test is indicative

of) than is the 1020.

In all cases, the fully martensitic structures show higher

impact values than any of the duplex structures, presumably due to

the effects of the second phase in stress intensification. Predicting

behavior from 50 to 100%. martensite may be considered purely spec­

ulative, as it would be below 25%. Impact values for a completely

ferritic structure, while having practical significance, would be

useful in showing the relationship here. Isolated tests indicate

that the impact energies absorbed by the completely ferritic structure

exceed those of the martensitic structure,

b. DBTT (1020).

Fullsize 1020 specimens were tested from -75 to +90 °C to

determine the ductile-brittle transition temperature (DBTT) for the

 24

fully martensitic and duplex steels. As is evident from Figure 20

r
~' there" is no sharp transitions; approximately 150 °C separates the
1
*'upper" and lower shelf energies. Using the average energy criterion,

•*the'DBTT would be approximately 25 °C; room temperature impact values

are about 1/2 of the upper shelf energy. An important observation

* ' is* that no significant change in the shape of the DBTT curves occurs

as V varies. SEM observations of the surfaces fractured at 90 °C

m
revealed thatthe fully martensitic structure failed in a completely

ductile manner, while the duplex alloys showed some evidence of

cleavage. This indicates some change in DBTT from the fully martens­

itic to the duplex alloys, but the difference is small. Of interest

also is that at 90 "C the values for 25% martensite are very close

to the 100% martensite values with the consideration that some cleavage

was still apparent in A, the true upper shelf values are expected to
be approximately equal.
- i - i

' " i ' ' C. MECHANICAL PROPERTY/MICROSTRUCTURE CORRELATIONS

, ' jfv.1' I
1. Quantitative (Empirical) Relationships.
An initial objective of this research was to determine if the
•j ,- i

-behavior of the duplex alloys could be approximated by the well known

,•' •% i v • - *'
1
rule for two phase mixtures:
r

S = S V + S,(l - V ) (1)
c m m i m
 25

where S = ultimate strength of the duplex alloy

S = ultimate strength of the martensite phase

m
S. = stress in the ferrite at the ultimate tensile

strain of the martensite

V - volume fraction of the martensite.

m

It is useful to rearrange the equation to facilitate correspond­

ence to the graphical results as:

S ( S S V + S la
c - m - f> m f ' < >

In this form, (S - S,) refers to the slope of the line, and

S. is the intercept at 0% martensite. The ultimate strength of the

martensite is known to be strongly influenced by carbon content, and

many researchers have derived equations relating strength to carbon

content ' by such methods as least squares analysis. It is gen-

1/3
erally accepted that S is proportional to C . The tensile strength

of the ferrite is assumed to be constant for the following reasons:

a) the carbon content of the ferrite is essentially constant, and

b) the major strengthening mechanism of the ferrite, subgrain forma­

tion, is beleived to be constant also. Analysis of numerous transmis­

sion electron micrographs showed no significant differences in subgrain

size from the intermediately quenched 1010 and 1020 steels, or at

varying martensite percentages. Microhardness readings in the ferrite

areas of numerous fractured tensile specimens were almost identical,

and substantiate this result.

 26

Equation 1 is derived for the conditions that the Strong phase

particles (martensite in this case) are unidirectionally aligned, and

equally strained. These are'clearly not the conditions in the duplex

alloy, and any attempt to quantitatively relate to the equation would

be empirical only.

The results of the tensile tests are shown in Figures 15 and

16, and assume an almost linear relation over the range of martensite

fractions tested. The fact that this is not Equation 1 exactly can

'"" be shown easily. If (S - S,) is the slope of the line, it Is constant.

m l
From previous discussions it is known that S varies significantly

with carbon content while S is unchanged. £ Thus for the duplex system

there is actually a small amount of curvature.: The slope of the 1020

ultimate strength curve is slightly greater than for the 1010 curve,

which reflects the higher strength of che 1020 martensite at equal

volume fractions (carbon content is higher). Thus Equation 1 appears

to be sensitive to the duplex system in this respect.

A serious omission from Equation 1 can be determined by

examination of Figure 15 (1010 steel) with special attention to the

ultimate strength curves. The large difference in the slopes (S^ - S ) f

J
- * > " «-i '• ' fi •'•:'. " '•''".•'' '
I can not be attributed to S because both steels are of the same
j i,..; ' , - > - „ ' - . , i'. • f P • - '
, overall composition, so at equal volume fractions, the carbon contents
of the martensites are the same and the strengths are the same.
;'••: »l <i , - • 5 ' . >•• ' V"; ' '• '-

From the previous discussion S, i s unchanged. Having eliminated or

accounted for these aspects, attention i s turned to the microstructural
differences in the two. Examination by transmission electron micros-
 27

copy has shown no morphological differences in the martensitic phases

of the duplex structures obtained by continuous annealing or interme­

diate quenching. Some of the difference in slope may be assigned to

the difference in the ferritic areas: the intermediately quenched

material had a ferrite subgrain structure, whereas no observable

division of the continuously annealed ferrite into subg'rains occured.

Extrapolation of the curves to 0% martensite shows a 42 ksi difference

in S , which can. not be totally attributable to the subgrain structure

f

difference. The obvious dissimilarity is illustrated in the optical

micrographs of Figure 3. C is from specimen CO (continuously annealed

1010) and D is from specimen IB (intermediately quenched 1010). The

shape of the martensitic regions [from the two processes] is drastically

different, and becomes suspect as the cause for the different behaviors.

Thus Equation 1 is inadequate and needs to be modified to include

factors concerning the detailed morphology [size, shape and distribu­

tion] of the second phase in the duplex structure. This result differs
34
from that found for duplex austenitic-martensitic steels.

Tempering changes both S and S , and it is desirable to

f

isolate, as much as possible, the effect of S .

f From Table 3 the

change in S_ from the as quenched to the 200 °C tempered condition

can be calculated for the fully marte&Bitic specimens; at 0.1% C the

change is 1% and at 0.19% C the change is 4%. Assuming a linear

change, which is at least qualitatively reasonable it is possible to

predict the tempering responses of higher carbon martensites (with

0.5% Mn) at 200 °C. A complicating factor that remains is that the
 28

carbon- content in the martensite cannot be determined"accurately;

t Bain and Faxton -and'Irvine have'related Hardness values to carbon
> content of marterisites containing up to 0.5% carbon,' arid the converted
microhardness -readings from the as quenched 1010 and 1020 martensites
are-in good agreement with the published figures. If a curve using
- the hardness' values obtained for 1010 and 1020 is plotted parallel
.-to-that-of Bain and'Faxton, and the relationships of Figure 21A (upper
J
- drawing) are used, the carbon content in the'martensites bf the 1010
i duplex structure'can be estimated. The values obtained from such
an estimation are tabulated below Figure 21. A rough, qualitative
.approximation of the effect of S on the strength of the duplex system
f

-becomes possible by making a comparison to a formula relating carbon

> to tensile strength, e.g., reference 15, and assuming a. linear response
to tempering at 200 °C.
Using the results obtained from the estimations and inserting
' -the values (as relative magnitudes) in Equation 1 shows that for the
tempered'condition, the stress in the ferrite, S , decreases as the
f

volume fraction of the ferrite decreases. Comparing the relative

magnitudes of the contributing factors,i.et, iS_V_ and S_(l - V ^ ) ,
ascertains'the,validity of the following: a s V . increases, the carbon
.content in the martensite decreases, therefore the ultimate strength
of"the martensite "also decreases. However, the total martensite
contribution to the duplex strength, S , increases. The stress in the
ferrite decreases as the martensite fraction increases; the total
contribution of the ferrite to the duplex'strength also decreases.
 29

It is emphasized that with so many approximations and assump­

tions that only a very limited estimate of the behavior is possible.

Nevertheless, it can be seen that Equation 1 is a useful empirical

equation to describe the mechanical properties of these duplex systems.

2. Qualitative Relationships.

It is immediately obvious from Figure 23 that the strength/

ductility combinations of the intermediately quenched 1010 are superior

to those of the continuously annealed alloy (1010 C). This can be

attributed to the; finer dispersion of martensite in the fefrite matrix,

and to the ferrite subgrain strengthening. An advantage may also be

assigned to the fewer fibrous carbides at the ferrite/martensite

interfaces. With the relative inferiority of the continuously annealed

1010, the remainder of this correlation will concern only the interme­

diately quenched 1010 and 1020.

In both the 1010 and 1020 duplex steels, the network of sub-

grains was very obvious. Analysis of numerous transmission electron

micrographs and general observations proved that the average subgrain

spacing was approximately 0.5 microns for both steels, and appeared

to be essentially independent of V . Thus, the strengthening variance

due to subgrains is believed to be negligible.

There is much controversy in the literature concerning the

beneficial aspects of retained austenite. Its ability to effectively

inhibit crack growth has especially been contested. However, the

amounts of retained austenite detected in the duplex alloys was very

small, and the effect it would have on mechanical properties is

 30

« ilikewise expected to be very small. ;In.any casev;small amounts were

, detected in. both 1010 and 1020, and at all volume fractions of marten-

site observed, and any property difference attributable to retained

austenite is again expected to be negligible- -r ,

Similarly, the amount of autotempering was essentially

independent of V and was observed in approximately equal amounts

in 1010 and 1020. As mentioned previously, the prior austenitic

grain size for all duplex alloys was ASTM #5. No significant strength­

ening differentials could thus be attributed to these metallurgical

parameters.

While no deliberate effort was made to quantify the amounts

. of twinning in the duplex specimens, twinning seemed more prevalent

35
_ in specimen 2B (1020) than in specimen IB (1010). Kelly and Nutting

have shown that increased carbon and the subsequent lowering of M

increased.'the probability of transformation twinning; an assertion

36
•> corroborated by Thomas. The martensite in2B (V = 40%) is expected
m
--, to be of considerably higher carbon content than that in IB (V = 35%),
?
' and the limited experimental observations are thus supportive of the

-. previous arguments. The effect of twinning on martensite toughness

has been fairly well documented; for as quenched steels with the same

<** carbon contents, fracture toughness-decreases with increasing twinning.

"•ii'.Twinned^martensites In carbon containing steels have also been shown

to i have inferior fracture toughness compared to untwinned martensites
- 38
at the same;strength level. A comparison of the subsize charpy
values VSJ yield strength (Figure 22) shows the duplex 1020 to have
 31

impact resistance which is considerably inferior to the duplex 1010.

This is presumably due to the larger amounts of twinning in addition

to the higher carbon content in the 1020.

The final variable to be considered appears to be the most

influential. For a given steel, the volume fraction of martensite

(and its carbon content) is seen to be the major controlling factor

for the strength, ductility and impact resistance of the duplex mate­

rial. Examination of Figures 15-17 and 19 supports this conclusion.

With the other metallurgical variables equal, the higher carbon content

in the 1020 martensite vs. the 1010 martensite at equal volume frac­

tions is chiefly responsible for the differences in mechanical prop­

erties .

D. COMPARISON TO COMMERCIAL SPECIFICATIONS

Figure 23 shows the strength/ductility combinations obtained

with the intermediately quenched, duplex 1010 and 1020 alloys in this

investigation. Also shown are the tensile properties of the commercial

1010 and 1020 in the hot rolled and cold drawn conditions.

Extrapolation of the curves shows that the properties obtained

with the duplex steels meet or exceed the commercial specifications

(compose the as quenched to the cold drawn, and the tempered to the

hot rolled), without the need for thermomechanical treatment. The

strengths obtained in this investigation greatly exceed the strengths

of commercial 1010 and 1020, with the consequent sacrifice in ductility.

As mentioned previously, the final selection of properties is determined

mmmmmmsmmmm

32

by the application for, which t^ fteel: wi.li fee-used,,,,;.

r ;

, M There are no conmercial fpeci£ications: available for 1010 and

:

1020 impact values, and typical values could not; be found in the

literature for comparison.

;

j - 'S-j,'.i .

^ ,.. i t^tKfrz?;^
 33

IV. CONCLUSIONS

Based on the Investigations and characterizations of the fully

martensltlc and duplex 1010 and 1020 steels In this research, the

following conclusions are made:

1. The rule for two phase mixtures, S = S V + S.(l - V ) , may

r
c m m f m *

be used empirically to describe the mechanical properties of

these duplex alloys. Suitable modifications to Include the

size, shape and distribution of the second phase are necessary

for more quantitative correlations.

2. For the 1010 steel, the Intermediate quench route enables

strength/ductility combinations to be obtained that are supe­

rior to those obtained by continuous annealing.

3. The largest factor controlling the strength, ductility and

impact resistance of the duplex alloys in this investigation

is the volume fraction of martensite.

4. The property differentials between the duplex 1010 and 1020

(intermediately quenched) at equal volume fractions of marten-

site are due mainly to the higher carbon content (hence,

higher strength) of the 1020 martensite.

5. Using the average energy criterion, the ductile-brittle transi­

tion temperature for the fully martensiticJ and duplex 1020

steels is about 25 °C. The DBTT appeared to be essentially

independent of V .
 34

It is suggested that the poorer impact properties of the

duplex 1020 steel (compared to 1010) are attributable to

the large amounts of transformation twinning and higher

carbon content in the 1020.

The tensile properties of the duplex 1010 and 1020 steels

in this investigation meet or exceed commercial specifications,

without the need for thermomechanical treatment.

 35

REFERENCES

1. Technology Forecast, Metal Progress, p. 32, Jan. (1977).

2. J. Y. Koo and G. Thomas, Mat. Sci. and Eng., 24, 187 (1976).

3. J. Y. Koo and 6. Thomas, Met. Trans., 8A, 525 (1977).

4. J. Y. Koo, M.S. Thesis, University of California, Berkeley,

LBL-3587, Feb. (1975).

5. S. Hayami and T. Furukawa, Micro Alloying '75, Symposium on

High Strength, Low Alloy Steels, Products and Processes, p. 78,

Washington, D.C. (1975).

6. R. L. Cairns and J. A. Charles, J. Iron and Steel Inst., p. 1044

(1967).

7. M. Enomoto and E. Furubayashi, Mat. Sci. and Eng., 24, 123 (1976).

8.. S. K. Huang, M.S. Thesis, University of California, Berkeley,

LBL-3503, Dec. (1974).

9. S. Jin, S. K. Hwang and J. W. Morris, Jr., Met. Trans., 6A, 1721

(1975).

10. R. L. Miller, Met. Trans., .3, 905 (1972).

11. E. Snape and N. L. Church, J. of Metals, p. 23, Jan. (1972).

12. R. A. Grange, Met. Trans., £, 65 (1971).

13. Automotive Engineering, 84, 49, Nov. (1976).

14. Edgar C. Bain and Harold W. Pax ton, Alloying Elements in Steel,

American Society For Metals, Metals Park (1966).

15. W. H. McFarland, Trans. TMS-AIME, 233, 2028 (1965).

16. K. J. Irvine, J. Iron and Steel'Inst., p. 820 (1962).

 36

17. Annual Book of ASTM Standard, designation E23-72, sections 8.2.1

and 8.2.2, revised June.(1972),

18. Robert DeHoff and Frederick Rhines, Quantitative Microscopy,

' McGraw-Hill, New York (1968);

19. M. Hansen, Constitution of Binary Alloys, 2nd Ed., McGraw-Hill,

New York (1958).

20. G. R. Speich and H. Warlimont, J. Iron and Steel Inst., p. 385

(1968).

21. A. S. Tetelman and A. J. McEvily, Jr., Fracture of Structural

Materials, John Wiley and Sons, New York (1967).

1 0
22. G. R. Speich and W. C. Leslie; Met. Trans., 2» * 3 (1972).

23. Paul G. Shewmon, Transformations in Metals, McGraw-Hill,

New York (1969).

24. H. I. Aaronson, H. A. Domian and G. M. Pound: Trans. AIME, 236,

753 (1966).

25. H. W. Paxton, "Transformation and Hardenability in Steels",

Climax Molybdenum Co., Ann Arbor, Mich., p. 3, 1967.

26. D. H. Warrington, J. Iron and Steel Inst., p. 610, July (1963).

27. K. Campbell and R. W. K. Hoheycbmbe, Metal Science, 8_, 197 (1974).

28. G. Krauss and A. R. Marder, Met; Trans., £» 2343 (1971).

29. S. K. Das and G. Thomas, Met. Trans., I, 325 (1970).

30. C. A. Apple, R. N. Caron and G. Krauss, Met. Trans., 5_, 593 (1974).

31. B. V. N. Rao, J. Y. Koo and G. Thomas, Proc. 33rd Electron

Microscopy Soc. Am., Claitors, Baton Rouge (1975), p. 30.

32. J. Y. Koo, Ph.D. Thesis, in progress. •

N. J. Petch, Progr. in Metal Phys., 5_, 1 (1954).

P. R. Manganon and G. Thomas, Met. Trans. 1, 1587-1594 (1970).

P. M. Kelly and J. Nutting, Proc. Royal Soc., A259, 45 (1960).

G. Thomas, Met. Trans., 2, 2373 (1971).

J. McMahon and G. Thomas, 3rd Int. Conf. on Strength of Metals

and Alloys, Cambridge, England, 1, 180 (1973).

S. K. Das and G. Thomas, Trans. ASM, 62, 659 (1969).

 Table Z. Alloy compositions determined after austenitization (wt. %)

Alloy 1 (1010) Carbon 0.10 Manganese 0.5

Alloy 2 (1020) Carbon 0.19 Manganese 0.5

Table II. Mechanical Properties (as quenched)

Room Temperature'
Two Phase Ultimate Strength Charpy » ,
Designa- Temperature v Yield Strength e^ e, u Energy . (Ft-Lbs)
Treatment tion (°C) ksi (MPa) ksi (MPa) {%) (%> fullsize gubsize
1010 A+TP 1A 775 30 64 (442) 99 (683) 10 15 11
A+IP IB 795 35 66 (455) 104 (718) 9.5 14 11
A+TP 1C 815 45 79 (545) 113 (780) 8.5 13 8
A 100 124 (856) 144 (994) 4 8 15
A+CIP
m
CA 765 30 68 (470) 110 (760) 11
A+CTP CB 775 40 69 (477) 108 (746) 10
A+CTP cc 795 55 76 (525) 111 (767) 8
A+CTP CD 810 60 80 (552) 116 (801) 7
A+CTP CE 740 15 68 (470) 105 (725) 10 13
1020 A+TP 2A 725 25 57 (393) 96 (662) l!6 14 2.5
A+TP 2B 740 40 73 (504) 122 (842) • 5 12 n 1.5
A+TP 2C 755 50 75 (518) 127 (876) 7.5 11.5 8.5 :
I -
A 2M 100 170 (1173) 203 (1401) 4 8 . 6 8.5
26
A - Austenization at 1100 °C, agitated iced brine quench.
A+TP - Austenization followed by two phase annealing with intermediate quenching.
A+CTP - Austenization followed by continuous two phase annealing.
 39

Table I I I . Mechanical properties (tempered 30 min. at 200 °C)

Designation Yield Strength Ultimate Strength

ksi (MPa)
e
u V
ksi (MPa) (%) (%)

1010 1A2 51 (352) 80 (552) 12 21

. 1B2 53 (366) 83 (573) 10 "17

1C2 54 (373) 84 (580) 8.5 16

1M2 126 (869) 143 (987) 5.5 9

1020 2A2 50 (345) 77 (531) 13 19

2B2 58 (400) 95 (656) 11.5 15

2C2 65 (449) 104 (718) 10 13

2M2 164 (1132) 195 (1346) 5 9

 40

Table IV. Mechanical properties after tempering at 200, 400 and 600 "C

.Tempering ' *
e
Temperature Yield Strength Ultimate Strength u t
CO ksl (MPa) ksl (MPa) (%) (%)

1010 As Quenched > 66 ' (455) 104 * (718) 9.5 14

(V'=35)
- ' 200 53 (366) 83 (573) 10 17

400 52 (359) < 74 (510) 11 18

1
600 47 (324) 63 (434) 12 19
1020 As Quenched 73 (504) 122 (842) 8.5 12
(V -40)
200 58 (400) 95 (656) 11.5 15

400 56 (386) 82 (566) 16 25

600 55 (379) 71 (490) 18 27

 41

Table V. Impact t e s t i n g of f u l l s i z e 1020 for transition temperature

Temp erature (C)

-75 -50 -25 0 23 90

2A 4 11 18 21 35
2B
1
0
4

3.5
5

4
5.5

4
8

6
18

16
g. 2C
IB
2M 11 11 14 20 26 36
 42

FIGURE CAPTIONS

Schematics of heat treatments, (a) and (b) intermediate quench­

ing, (c) continuous' annealing (1010 only).

Specimen geometries for mechanical testing specimens.

Optical micrographs. A, starting microstructure (1010 and 1020

were identical): B, 1020 duplex-intermediate quenching. C,

1010 duplex, continuous annealing. D, 1010 duplex, intermediate

quenching.

Composite transmission electron micrograph of martensite lath

embedded in the ferrite of specimen 2B (1020). Note the

subgrain structure in ferrite.

Transmission micrographs of specimen 2B. A, martensite and

dislocated ferrite. B, subgrain network. C, dislocations

rearranged to form a subboundary. Some background precipitation

(not identical) appears in some areas.

Transmission micrographs. Fibrous carbides at ferrite/marten-

site interfaces. A, bright field of specimen cc (continuously

annealed 1010), B is the corresponding dark field. C, specimen

1C, intermediately quenched 1010.

Scanning electron fractographs of tensile failures of interme­

diately quenched specimens. All specimens showed a predominantly

ductile failure. A and B, as quenched. C, tempered at 200 "C

for 30 minutes.
 43

Scanning electron fractographs at increasing magnftieations

continuously annealed 1010. , Localized dual modes of failure

are evident as shown in C and D.

Transmission electron micrographs. A, composite showing broad

face of martensite lath. Note pocket boundaries and auto-

tempering. B, typical lath martensite in 1010 and 1020 steels.

C, complex intersection of laths in martensitic phase of duplex

specimen.

Transmission micrographs. Extensive autotempering in completely

martensitic structures of 1010 and 1020., steels.

Transmission micrographs of retained austenite in specimen 2M

(1020). A, bright field. B, corresponding dark.field where

thin films of retained austenite reversed contrast when (002)

reflection was imaged. C, selected area diffraction pattern.

D, indexed SAD proving K-S relationship of martensite and

austenite.

Transmission micrographs of specimen 2B (1020). A, dislocated

ferrite and ferrite/martensite interface. Note twins at upper

right. B,extremely fine transformation twins.

SEM fractographs of charpy failures at room temperature. Most

failures may be characterized by quasi cleavage. Note: only

intermediately quenched specimens were tested in impact.

SEM fractographs of charpy failures. A and B, room temperature

fractures. C, secondary cracking in 2B at -50 °C.

 44

* 15. Strength"and elongation ductility vs. marterisite fraction of

intermediately quenched and continuously annealed 1010. Open

and closed figures delineate yield and tensile strengths, unform

and total elongations. Note: No continuously annealed material

was tempered.

16.' Strengths and elongation ductility vs. martensite fraction for

intermediately quenched 1010 and 1020 steels. As quenched.

17. Strength and elongation ductility vs. martensite fraction for

intermediately quenched 1010 and 1020 steels tempered at 200 °C

for 30 minutes.

18. As quenched and tempered properties of intermediately quenched

1010 and 1020 steels containing 35% and 40% martensite,

respectively.

19. Charpy impact energies of intermediately quenched 1010 (subsize)

and 1020 (subsize and fullsize) steels.

20. Ductile-brittle transition curves for fully martensitic and

- duplex' 1020 steels.

"21.* 'Plot of mlcrohardneBS readings (vs. % martensite) in the marten­

sitic phase of continuously annealed duplex specimens. Approximate

-<< 'carbon contents vs. martensite fraction are tabulated below

the figure.

22. Charpy impact energies vs. yield strength for Intermediately

l
~' - ' ~ quenched 1010 and 1020 steels.
 45

23. Total elongations v s . ultimate strengths of the 1010 and 1020

s t e e l s i n t h i s investigations. Also shown are the minimum

commercial specifications of hot rolled and cold drawn 1010 and

1020 s t e e l s .
 46
3 0 m i n
M00°C|
Fe RICH PORTION
OF Fe-C PHASE
900 DIAGRAM

800

P 900
uj
'••3

&
tE
LI 800
Q.
2
UJ

0.1 0.2
WT%C
I lOOt 30min
900

800

WT%C
XBL775-5407

Fig. 1