influence of annealing treatment on the formation

7212019 Influence of Annealing Treatment on the Formation

httpslidepdfcomreaderfullinfluence-of-annealing-treatment-on-the-formation 114

Influence of Annealing Treatment on the Formationof NanoSubmicron Grain Size AISI 301 AusteniticStainless Steels

DL JOHANNSEN A KYROLAINEN and PJ FERREIRA

Nanosubmicron austenitic stainless steels have attracted increasing attention over the past few years

due to fine structural control for tailoring engineering properties At the nanosubmicron grain scalesgrain boundary strengthening can be significant while ductility remains attractive To achieve a nano submicron grain size metastable austenitic stainless steels are heavily cold-worked and annealed toconvert the deformation-induced martensite formed during cold rolling into austenite The amountof reverted austenite is a function of annealing temperature In this work an AISI 301 metastableaustenitic stainless steel is 90 pct cold-rolled and subsequently annealed at temperatures varying from600 degC to 900 degC for a dwelling time of 30 minutes The effects of annealing on the microstructureaverage austenite grain size martensite-to-austenite ratio and carbide formation are determinedAnalysis of the as-cold-rolled microstructure reveals that a 90 pct cold reduction produces a combi-nation of lath type and dislocation cell-type martensitic structure For the annealed samples theaverage austenite grain size increases from 028 mm at 600 degC to 585 mm at 900 degC On the otherhand the amount of reverted austenite exhibits a maximum at 750 degC where austenite grains with anaverage grain size of 17 mm compose approximately 95 pct of the microstructure Annealing temper-atures above 750 degC show an increase in the amount of martensite Upon annealing (Fe Cr Mo) 23C6

carbides form within the grains and at the grain boundaries

I INTRODUCTION

AUSTENITIC stainless steels (SS) are frequentlyselected in applications where good corrosion propertiesand aesthetic considerations are important However incases where an austenitic SS sheet needs to be subjected tocold forming such as deep-drawing or stamping it exhibitsa large variation in yield strength (above 50 pct) and unde-

sirable yield strength in regions absent of deformation Themain cause of this variation is related to the cold-workingprocess during which the deformed regions transform fromthe soft parent face-centered-cubic ( fcc) austenite (g ) phaseinto the hard martensite (a9) phase In low-carbon auste-nitic SS the tetragonality of the martensite is basicallyabsent and thus the martensite phase can be consideredas exhibiting a body-centered-cubic (bcc) structure

A feasible solution to resolve the variation in yield strengthassociated with the cold-working process has not beenachieved mainly because in the forming processes deforma-tion and hard martensite formation occur in areas where highstrength is not required while regions where high strength isnecessary are not subjected to deformation leaving soft aus-tenitic regions In these soft areas the material can be easilyscratched dented and deformed in service and thus its cor-rosion properties and appearance strongly deteriorate

A tempting solution to this problem is to use a fullymartensitic SS for the forming operation instead of an aus-

tenitic SS In this fashion the undeformed regions maintainthe high strength and hardness of the martensite phasewhereas the deformed regions acquire the high strengthand hardness associated with the martensite phase How-ever due to the poor formability of the martensitic struc-ture products manufactured in this fashion may fracture

To overcome the aforementioned limitations a researchplan has been designed to develop in cold-rolled SS sheets

regions (10 to 20 mm) with nanosubmicron austenite grainsize exhibiting high strength and high formability proper-ties while the sheet remains in the martensite phase out-side the narrow regions Using this concept a single SSsheet can be produced such that undeformed regions main-tain the strength and hardness of the martensite whiledeformed regions have enhanced strength and hardnessdue to the refined austenite grain structure produced withinthe narrow regions (Figure 1) In particular the idea con-sists of heavily cold-rolling an AISI 301 austenitic SS sheet(90 pct reduction in thickness) to induce the formation of martensite (Figure 1 step 1) followed by a local annealing

treatment within the 600 degC to 900 degC temperature range tonucleate and reform the austenite phase in narrow specificregions (Figure 1 step 2) before subjecting the samples tocold forming (Figure 1 step 3) Grain sizes achieved by thismethod are below 1 mm[1ndash6] In this manner it is possible tolocally control the nanosubmicron grain size and conse-quently the strength of the material before forming

Within the wider scope of the aforementioned idea theinitial research objective described in this paper is to iden-tify during step 2 the influence of annealing temperatureon the microstructure of a metastable austenitic AISI 301SS Typical factors affecting the reversion from martensiteto austenite are the chemical composition amount of coldworking annealing temperature and dwelling time Althoughprevious experiments on Fe-Cr-Ni steels and plain carbon

DL JOHANNSEN (now DL Rittermann) formerly a GraduateResearch Assistant in the Materials Science and Engineering Program atthe University of Texas at Austin Austin TX 78712 is now with IntelOregon A KYROLAINEN Senior Researcher is with Research Labora-tories Outokumpu Stainless Steel Oy Tornio Finland PJ FERREIRAAssistant Professor is with the Materials Science and Engineering Pro-

gram University of Texas at Austin Austin TX 78712 Contact e-mailferreiramailutexasedu

Manuscript submitted October 29 2005

METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 37A AUGUST 2006mdash2325



steels have shown correlations between amount of coldwork annealing temperature and dwelling time on themicrostructure achieved very limited work has been doneon commercial austenitic SS alloys containing carbon andnitrogen[127] In addition most of the research performedso far has dealt with the effects of annealing on the rever-sion mechanisms but a strong correlation between the grainsize and the microstructure achieved has not been investi-gated In particular a fundamental understanding of theeffects of the annealing temperature on the microstructuremartensite-to-austenite (a9 g ) ratio austenite reversion andaverage grain size has not yet been accomplished

The goal of this paper is thus to correlate the annealingtemperature with the reverted austenite grain size phasefraction and microstructure achieved in a commercial AISI301 SS via magnetic measurements X-ray diffractometry

and transmission electron microscopy (TEM) experimentsA commercial AISI 301 metastable SS grade was selectedbecause its metastable austenitic structure facilitates theformation of stress-induced martensite and subsequentlypromotes an efficient austenite reversion In addition AISI301 SS grades are commercially available thereby widen-ing the potential impact of this research work

II EXPERIMENTAL PROCEDURE

A Materials

The AISI 301 SS used in this work was provided byOutokumpu Stainless Oy Finland with an alloy composi-tion (wt pct) given in Table I This SS has an Md30 tem-

perature of 18 degC[89] and a calculated MS temperature of ndash118 degC[1011] The material was produced by continuouscasting and subjected to hot rolling heat treatment and afinal 90 pct cold-rolling reduction at Outokumpu StainlessOy The amount of retained d-ferrite was found to be appro-ximately 02 vol pct As-received 90 pct cold-rolled sheetswere cut into rectangular specimens subsequently subjectedto a heating rate of 100 degCmin and annealed isothermallyfor 30 minutes at the temperatures 600 degC 650 degC 700 degC750 degC 800 degC 850 degC and 900 degC followed by forcedair-cooling For comparison one specimen was left in theas-cold-rolled (CR) state

B Methods

1 X-ray diffractionRectangular sections approximately 5 3 7 3 08 mm3

were cut with the rolling direction (RD) parallel to the longdimension These samples were subsequently measured in aPhillips PW1720 X-Ray Diffractometer (XRD) operating ata voltage of 40 keV and a current of 40 mA The tests wereperformed at ambient temperature using CuK a1 radiationPeak measurements were taken from 30 to 90 deg (2u) insteps of 01 deg with a dwell time of 4 seconds From theresulting intensity vs 2u plots MDI Jade 65 software was

used to index the peaks The results were compared topublished XRD data for an austenitic SS[12]

In addition the X-ray spectrum of the CR specimen wasused to calculate the volume fraction of austenite and mar-tensite in this sample based on the method published byDickson[13] According to this method the volume fractionsof austenite Cg and martensite Ca were evaluated on thebasis of the first three reflexions for the austenite ((111)(200) (220)) and the martensite phase ((110) (200) (211))in the form[13]

Cg

Ca

5

1

ng

+ ng

0

Ig

Rg

1

na+na

0

Ia

Ra

[1]

where Ca frac141= 11ethCg =CaTHORN

Cg frac14ethCg =CaTHORN= 11ethCg =Ca

na and ng are the number of martensite and austenite reflex-ions considered Ia and Ig are the integrated intensities of martensite and austenite for the reflexions considered andRa and Rg are the relative intensities of martensite andaustenite for the reflexions considered On the basis of Eq [1] the volume fraction of martensite present in theCR (90 pct reduction in thickness) sample can be calculatedas 974 pct This sample was selected as the reference valuefor the determination of phase fraction in all the samples by

Fig 1mdashNovel method used to produce nanosubmicron-grained austenitewith good formability properties

Table I Chemical Composition of AISI 301 Stainless Steel(Weight Percent)

C Cr Ni N Mn Si Mo

0096 167 66 00635 123 118 072

P Cu Co S D Y Fe

026 017 011 0001 00031 0005 Balance

2326mdashVOLUME 37A AUGUST 2006 METALLURGICAL AND MATERIALS TRANSACTIONS A



correlating The obtained X-ray results with The SQUIDmagnetic measurements described below

2 SQUIDTo determine the fractional amount of martensite (ferro-

magnetic phase) and austenite (paramagnetic phase) eachsample was measured in a Quantum Designrsquos Supercon-ducting Quantum Interference Device (SQUID) This sys-tem uses the extreme sensitivity of the SQUID to detectminute changes as small as 1 3 1014 T in a specimenrsquos

response to an applied field Due to its high sensitivitymeasurement errors are typically below 001 pct For theSQUID tests specimens with dimensions approximately1 3 1 3 5 mm were cut with the long dimension parallelto the rolling direction and polished to remove irregularedges and surface oxidation The magnetic field was appliedparallel to the rolling direction of the specimen The testswere performed at 27 degC in increasing steps of 00250 Teslaup to a magnetic field of 02 T and then in increasing stepsof 01 T until a field of 1 T was reached Each data point wasmeasured five times and averaged The magnetic response of the stainless steel as a function of the applied magnetic field

was measured and normalized with respect to the specimenvolume A maximum applied field of 1 T was chosen to ensurethe saturation of the magnetic phases present in the alloy

To determine phase fraction and to establish a firm con-fidence in the SQUID measurements a reference value wasselected based on the volume fraction of martensite presentin the CR (90 pct reduction in thickness) sample whichwas calculated by X-ray as 974 pct (see Section IIndashBndash1)As the measured magnetization per unit volume of this CRsample is 633 emucm3 the saturation magnetization perunit volume mSa91T of a fully 100 pct martensitic structurecan be calculated as approximately 650 emucm3 if weassume a linear behavior However as the CR sample

exhibited 02 volume percent of d-ferrite (d has a saturationmagnetization of 971 emucm3[14]) we need to subtract thecontribution of d-ferrite from the measured 677 emucm3Under these conditions a CR sample containing 100 pctmartensite should correspond to a saturation magnetizationof 653 emucm3

3 Transmission electron microscopyA JEOL 200CX and a JEOL 2010 TEM operating at 200

kV and equipped with single and double tilt stages were usedto identify the overall microstructure austenite grain sizeand dislocation substructure For each annealing condition

three samples were observed Representative images weretaken at each annealing temperature The JEOL 2010 TEMis equipped with an Oxford energy dispersive spectroscopy(EDS) system which was used for chemical compositionanalysis of second-phase particles present in some of thesamples EDS spectra were also collected from areas adjacentand away from the second-phase particles for comparison

Electron transparent disks were made using a StruersTenuPol-5 Twin-Jet electro-polisher using the followingelectrolyte and parameters 590 mL m-Butanol 350 mLmethanol and 60 mL perchloric acid performed at ndash10 degCat a voltage of 2530 V and a current of 24 mA

Austenite grain sizes were calculated from scanned TEMnegatives using standard ASTM E112 The average grainsizes were calculated from a measurement of 100 grains

4 Hardness measurementsHardness measurements were performed in the CR and

annealed samples using the Rockwell B hardness testmethod In this technique a hardened steel ball indenteris forced into the test material under a preliminary minorload of 10 kgf and while the preliminary minor load is stillapplied the sample is subjected to an additional load of 100 kgf Ten measurements were obtained from each sampleand an average value was calculated Hardness Rockwell Bnumbers (HRB) were subsequently converted to hardness

Vickers numbers (HV) using the ASTM E140 standardFinally HV values were converted to yield strength valuesusing the linear relation HV 5 3s y 98[15]

III RESULTS

A X-ray Diffractometry

Figure 2 shows the X-ray spectra for the CR sample andthe samples annealed at various temperatures For eachspectrum the intensity of each peak was normalized withrespect to the total spectrum intensity As shown in Figure2 the CR sample exhibits primarily martensite peaks with asmall (220) austenite peak As the annealing temperatureincreases from 600 degC to 800 degC the austenite peaksincrease in intensity with respect to the martensite peaksrevealing an increase in the volume fraction of austeniteThis trend is reversed above 800 degC for which the austen-ite-to-martensite peak ratios decrease indicating anincrease in the volume fraction of martensite

A comparison between the obtained spectra (Figure 2)and published results for randomly oriented polycrystallineaustenite and martensite allows us to determine qualitativelythe existence of preferred orientation This analysis showsthat the CR sample has a definitive preferred orientation

along the 211a9 planes of martensite and the 220g planes of austenite However upon annealing at 600 degC achange occurs in the martensite phase with the preferredorientation shifting toward the 110a9 planes This trend

Fig 2mdashXRD spectra for the cold-rolled (CR) and annealed samples atvarious annealing temperatures




becomes more evident above 650 degC (Figure 2) On the other

hand the austenite phase maintains the 220g preferredorientation up to 650 degC while at 700 degC and above theaustenite reversion becomes more pronounced and gradu-ally acquires a more random orientation (Figure 2)

For the determination of the volume fraction of austeniteand martensite X-rays were used and correlated withSQUID measurements in the CR sample to establish areference value On the basis of this reference sampleand due to the superior accuracy of the SQUID the volumefraction of austenite and martensite was calculated for allsamples using the latter technique (see next section)

B SQUID Magnetic Measurements

Because of the accuracy of the SQUID method magneticmeasurements were used to determine the volume fractionsof martensite and austenite The saturation magnetizationper unit volume of the CR specimen (no annealing) wasfound to be 633 emucm3 The measured saturation magneti-zation per unit volume for all annealed samples (measurederror around 6104) is shown in Figure 3

Using the saturation magnetization data of Figure 3 it isthus possible to calculate the volume fraction of martensiteand austenite as discussed in Section IIndashBndash2 The phasefractions for each annealing temperature converted from

the saturation magnetization data are shown in Figure 4The data show that the maximum austenite reversion occursfor an annealing temperature of 750 degC above this temper-ature the amount of martensite increases

C TEM Observations

1 Austenite grain sizeFigure 5 provides the average grain size diameter

obtained at each annealing temperature The average grainsize was calculated from measurements of 100 grains Theerror bars represent the deviation from the mean grain sizegiven a 95 pct confidence limit As shown in Figure 5 theaverage grain size starts to increase above 650 degC andsuffers a slight decrease at 800 degC above this temperature

a drastic increase in average grain size is evident Figure 5also shows that wider grain size variations are found above800 degC However a calculation of the ratio between thelargest and the smallest grain sizes found among 100 grainsshows that the most significant effect (ratio 40) is observed

at 750 degC all other temperatures exhibit a ratio below 20

2 Microstructural analysis

a CR specimenThe CR sample shows a microstructure composed of

two distinct martensite structures regions of lath martensiteand regions of dislocation cell-type martensite These tworegions can be clearly identified by the diffraction patternsobtained (Figure 6) the regions of lath martensite producea diffraction spot pattern whereas the regions of dislocationcell-type martensite produce a diffraction ring pattern Thering-like diffraction pattern exhibits brighter portions insome regions indicating that some preferred orientationis present For the purposes of grain refining the dislocation

Fig 3mdashSaturation magnetization as a function of annealing temperatureAll samples were annealed for 30 minutes at the respective temperatures

Fig 4mdashThe volume fraction of the martensite and austenite phasesdetermined by magnetic measurements

Fig 5mdashThe average austenite grain size as a function of annealingtemperature




cell-type martensite is ideal because it contains a muchhigher dislocation density than the lath-like martensitewhich can act as nucleation sites for the austenite duringannealing[61617]

Also important is the fact that some retained austeniteis present in the CR sample as shown by the diffractionpattern in Figure 6(b) which confirms the SQUID resultsobtained

b 600 degC annealing

Figure 7 shows the microstructure of the samples anne-aled at 600 degC at different magnifications In Figure 7(a)the cold-rolling direction is still readily apparent (whitearrow) Figure 7(b) shows a higher-magnification image of Figure 7(a) where the presence of austenite is visible This isconfirmed by Figure 7(c) which shows elongated grains of austenite (outlined in black) adjacent to regions of lath mar-tensite and by the electron diffraction spot pattern shown inFigure 7(d) taken from the white circled area in Figure 7(a)Many of the austenite grains have low dislocation densityassociated with the nucleation and growth of new crystals

A careful examination of the microstructure revealed anaverage austenite grain size of approximately 028 mm(Figure 7(c) shows an example of two austenite grainsdelineated by dark lines) and a narrow grain size variation

(Figure 5) which indicates that the sample is going throughthe very initial process of phase conversion Furthermoresecond-phase particles approximately 20 nm in size andidentified as carbides (see Section IIIndashCndash3 for a thoroughanalysis) were detected on large reverted austenite grains(black circle in Figure 7(c))

c 650 degC annealingThe microstructure of the samples annealed at 650 degC is

shown in Figure 8 Figures 8(a) through (c) show recrystal-lized ultrafine grains of austenite as well as signs of lath-type

Fig 6mdash(a) TEM micrograph showing the two distinct martensite struc-tures in the CR specimen (b) Diffraction spot pattern of the lath martensitetaken from the region represented by the white circle B5 frac12111a0 Resid-ual austenite is also identified in the spot pattern (c) Diffraction-ringpattern of the dislocation cell-type martensite taken from the region rep-resented by the black circle fcc spots corresponding to the austenite phaseare also shown

Fig 7mdash(a) Microstructure of the 600 degC sample The white arrow shows thecold rolling direction (b) Higher magnification of (a) showing recrystallizedfine-grained austenite (c) Higher magnification of (b) showing the presenceof lath martensite adjacent to austenite grains several of which are outlined

in black The presence of carbides can be seen within the small black circle(d ) Electron diffraction pattern taken from the circled area in (a) indicatingthe presence of austenite B 5 [001]g

and quasi-rings of the martensite




martensite The ultrafine scale of the austenite grains can beconfirmed by the still large number of spots configured in aring-like pattern (Figure 8(b)) whereas the presence of martensite is revealed by the spot diffraction pattern shownin Figure 8(c) Figure 8(d) shows a higher-magnificationimage of the austenite grain structure evident in Figure8(a) A careful examination revealed the presence of equi-axed grains with an average grain size of 028 mm Similarto the samples annealed at 600 degC the annealing treatmentat 650 degC produced a narrow grain size variation (Figure 5)

This result indicates that at 650 degC the conversion processfrom martensite to austenite is still underway Carbideswere also detected on some of the larger reverted austenitegrains (black circles in Figure 8(d))

d 700 degC annealingThe microstructure of the sample annealed at 700 degC

differed significantly from the samples annealed at lowertemperatures Figure 9 shows regions of large reverted aus-tenite grains (region L in Figure 9(a)) surrounded by pock-ets of small austenite grains (region S in Figure 9(a)) Thepresence of areas with fine-grained austenite can be con-firmed by the ring-like electron diffraction pattern shown in

the insert of Figure 9(b) As expected this sample exhibitsa large deviation in grain sizes from the measured 83 mmaverage grain size (Figure 5)

Further TEM observations (Figure 10) confirmed thepresence of martensite in the sample annealed at 700 degCas also detected via SQUID and XRD In Figure 10(a) aslip band separating two different regions of austenite isobserved In addition carbides were most noticeable onthe larger austenite grains

e 750 degC annealingSimilar to the sample annealed at 700 degC the micro-

structure of the 750 degC sample shows regions of largeand small austenite grains (Figure 11(a)) However asdepicted in Figure 11(a) fewer pockets of ultrafine-grainedaustenite and wider areas of larger austenite grains areobserved TEM observations revealed an average austenitegrain size of approximately 174 mm with a significantdeviation from the average (Figure 5) This large deviationfrom the average is a clear indication of the simultaneousphase conversion and grain growth processes

In addition as shown in Figure 11(a) a large number of carbides with an average diameter of 110 nm were presentin the 750 degC sample In Figure 11(a) the white arrowindicates strings of carbides consistent with precipitation

along previously existing martensitendashlath boundariesAs expected from the SQUID and X-ray results the

TEM observations confirmed the presence of some retainedmartensite as shown in Figure 11(b)

f 800 degC annealingAs shown in Figure 12 the microstructure of the samples

annealed at 800 degC looks very different from the samplesannealed at lower temperatures This apparent differenceis due mainly to a sharp increase in the number of defectspresent in the austenite grains Due to the presence of reverted austenite with a high density of defects the mar-tensite phase is visually more difficult to distinguish How-ever as shown in Figure 12 lath martensite can be identifiedconfirming the results from SQUID and XRD testing

A careful examination of the austenitic regions (Figure13) shows austenite grains with a high density of defectsand small equiaxed defect-free austenite subgrains (labeledsg in Figure 13) These subgrains are the byproduct of a shear-type reversion mechanism by which the reverted

Fig 8mdash(a) TEM microstructure of the 650 degC sample showing the newlynucleated austenite and lath-type martensite (b) Ring-like diffraction pat-tern of austenite grains taken from the region in (a) represented by theblack circle (c) Spot-like pattern of lath martensite taken from the regionin (a) represented by the white circle B5 frac12111a0 (d ) Higher magnifica-tion of (a) showing the presence of carbides




austenite microstructure maintains microstructural featuressimilar to the martensite

Although grain size measurements are very difficult toassess in the 800 degC sample a conservative approach indi-

cates an average austenite grain size of 126 mm (Figure 5)which is similar to the 750 degC sample However in contrastto the 750 degC sample the 800 degC sample displays a muchsmaller ratio between the largest and the smallest grainsizes Furthermore the 800 degC sample exhibits significantlyless carbide precipitation than the 750 degC sample althoughthe carbides present were approximately of the same size asthose found in the 750 degC sample

g 850 degC annealingThe microstructure of the samples annealed at 850 degC

shows the presence of a large number of defect-free equi-axed austenite grains and a high density of carbides (Figure14) When compared to the samples annealed at lower tem-peratures the samples at 850 degC exhibit a significant

increase in average grain size (55 mm) In addition thereare significantly fewer defects than in the 800 degC sampleThe carbides approximately 115 nm in size are distributedalong extended regions and are seen mainly in the center of the grains

Although martensite was not found within the aforemen-tioned austenite areas the samples annealed at 850 degC showthe presence of martensite (Figure 15) which confirm theresults obtained by SQUID and XRD

In addition carbides are visible within the lath-type mar-tensite as indicated by the small white circles in Figure15(a) As in the other samples the carbides are locatedwithin the grains and not at the grain boundaries

h 900 degC annealingThe microstructure of the samples annealed at 900 degC

show large reverted equiaxed austenite grains containingcarbides (Figure 16) The average austenite grain size wasmeasured as 585 mm although the ratio between the largestand the smallest grain sizes measured decreased withrespect to the 850 degC sample Moreover the austenitegrains are relatively free of defects when compared withthe samples annealed at 800 degC and 850 degC Figure 16 alsoshows a heterogeneous distribution of carbides with extended

Fig 9mdash(a) TEM micrograph of sample annealed at 700 degC for 30minutes showing large (L) and small (S) austenite grains (b) Highermagnification of (a) showing more clearly the variation in austenite grainsize The ring-like electron diffraction pattern shown in the insert andtaken from the region in (b) represented by the white circle confirms thepresence of ultrafine-grained austenite Carbides can be identified in themicrostructure within the black circles

Fig 10mdash(a) TEM micrograph of the sample annealed at 700 degC showingan area where martensite and austenite phases coexist A slip band (arrow)separates two regions of austenite with different grain sizes (b) Electrondiffraction pattern taken from the region in (a) circumscribed by the bot-tom left white circle B5 frac12111a0kB5 frac12011g (c) Electron diffraction pat-tern taken from the region in (a) circumscribed by the upper right white

circle B5

frac12111a0kB5

frac12011g




austenite areas free of carbides while other regions exhibitsevere carbide precipitation

The presence of martensite in this sample which was

detected by SQUID and XRD is also confirmed by TEMFigure 17(a) shows a region where both lath-type marten-site and reverted austenite can be seen As was the case inthe 850 degC sample the regions of lath-type martensite alsocontain the presence of carbides (the small white circle inFigure 17(b))

3 Second-phase particlesAs discussed in the previous sections second-phase par-

ticles were found in all annealed samples Upon annealingat 600 degC and 650 degC the number of second-phase particleswas relatively low and constant and their size were rela-tively small of the order of 20 nm (Figure 18) Annealingat temperatures between 750 degC and 900 degC led to a sig-nificant increase in the number of second-phase particles

formed and in the respective size of the particles The800 degC sample was an exception to this trend with virtuallyno second-phase particles visible (Figure 12) However forthe second-phase particles found the size of the existingparticles was similar to those present in samples annealedbetween 700 degC and 900 degC

Fig 11mdash(a) TEM micrograph of the sample annealed at 750 degC for 30minutes showing areas composed by small (S) grain sizes and large (L)grain sizes and severe carbide precipitation within the large grains Thewhite arrow indicates a string of carbides that nucleated at lathndashmartensiteboundaries (b) Region showing the presence of lath-type martensite in thesample annealed at 750 degC The electron diffraction pattern was taken fromthe circled region confirming the presence of martensite B5 frac12111a0

Fig 12mdash(a) TEM micrograph of the 800 degC sample showing revertedaustenite grains and lath martensite (b) Electron diffraction pattern of austenite taken from the region within the white circle ( c) Electron dif-fraction pattern of martensite taken from the region within the black circleB5 frac12111a0

Fig 13mdashMicrostructure of the austenite region for the sample annealed at800 degC showing the presence of defect-free subgrains (sg) and adjacentgrains with a high density of defects




The characterization of these second-phase particles wascarried out by EDS and TEM observations A comparisonof the EDS spectra from the matrix and the particles showsan increase in iron chromium and molybdenum contentin the second-phase particles (Figure 19) In addition acarbon peak can be identified in the EDS spectrum obtainedfrom the particles (Figure 19(b)) To conclusively identify

the second-phase particles TEM images and diffractionpatterns were obtained from the particles (Figure 20)Based on these observations the particlesrsquo crystal structurewas determined to be FCC with a lattice parameter of approximately 106 A These results in conjunction withthe EDS analysis confirm that the second-phase particlesare iron-chromium-molybdenum carbides of the typeM23C6 In practically all cases these observed carbidesexhibit the shape of globular particles (Figure 20) and areseen to precipitate predominantly in the austenite matrix

D Mechanical Properties

Rockwell B hardness measurements performed in the CRand annealed samples are shown in Table II Converted

Vickers hardness values and corresponding yield strengthvalues are shown in Figure 21 As depicted in Table II andFigure 21 samples annealed at 600 degC and 650 degC showvery high hardnessyield strength values As the annealingtemperature is increased from 650 degC to 750 degC there isa significant reduction in hardnessyield strength Above750 degC the hardnessyield strength values are essentially

Fig 14mdashTEM micrograph of the sample annealed at 850 degC showinglarge equiaxed grains of austenite and severe precipitation of carbides

Fig 15mdash(a) TEM micrograph of the sample annealed at 850 degC showingthe presence of martensite and carbides (within the small white circles) (b)Electron diffraction pattern of martensite taken from the region corre-sponding to the large white circle in (a) B5 frac12012a0

Fig 16mdashTEM micrograph of the sample annealed at 900 degC showinglarge reverted austenite grains and carbide precipitation

Fig 17mdash(a) TEM micrograph of sample annealed at 900 degC showing anarea where both austenite and martensite are present (b) Higher magnifi-cation of (a) revealing the carbides among the lath martensite (area withinthe small white circle) (c) Electron diffraction pattern taken from theregion corresponding to the white circle in (a) confirming the presenceof both austenite and martensite




very similar except for the sample annealed at 800 degCEven for samples annealed at 850 degC and 900 degC the yieldstrength is considerably higher than for conventional stain-less steels (Figure 21)

IV DISCUSSION

A Influence of Annealing Temperature on a9 g Ratio

As can be seen in the results from XRD (Figure 2)increasing the annealing temperature causes an increasein the amount of reverted austenite up to 800 degC Above800 degC the XRD data show an increase in the amount of martensite produced as depicted by an increase of the211a9 peak at 850 degC and 110a9 211a9 peaks at900 degC The maximum amount of reverted austenite shown

by XRD is thus different from the results obtained bythe SQUID method (Figures 3 and 4) In this case themaximum amount of reverted austenite achieved occursat 750 degC which is in agreement with the work byDi Schino et al[4] At annealing temperatures above750 degC an increase in the SQUID magnetic response isobserved which indicates an increase in the amount of martensite formed

This discrepancy between the XRD and the SQUID datais not yet fully understood However as shown in Figure14 the specimen annealed at 800 degC exhibits a very differ-ent microstructure from the samples annealed at 750 degC

(Figure 12) despite their similarity in terms of marten-site-to-austenite ratio As will be discussed later this sig-nificant difference in microstructure is attributed to achange in the mechanism for austenite reversion wherebyaustenite forms by shear reversion at 800 degC As a result itcan be argued that at 800 degC the texture produced byannealing could be rather different from the samplesannealed at 750 degC

One other interesting observation shown by both XRDand SQUID measurements is that the amount of revertedaustenite goes to a maximum and then decreases for anadditional increase in annealing temperature Although this

phenomenon has been reported in several works[256181920]

it has yet to be fully explained The argument that has beenpresented several times is that higher annealing tempera-tures cause an increase in the austenite grain size thusraising the MS temperature[256181920] As a consequenceupon cooling thermally induced martensite is more likelyto form for samples annealed at higher temperatures Whilethis statement may sound plausible the AISI 301 SS usedin this research has a calculated MS temperature of ndash118 degCfor an average grain size above 40 mm Since in this studythe largest reverted austenite grain size achieved is 58 mmat 900 degC we should then expect an MS temperature lower

than 118 deg

C based on the grain size argument

[256181920]

Thus for annealed samples cooled to room temperature itis unlikely for thermally induced martensite to form as aresult of larger austenite grains

A more likely argument for the increase in martensitecontent for annealing temperatures above 750 degC (assumingthe results obtained by the SQUID data) is that carbondepletion from the matrix due to the formation of the (FeCr Mo)23C6 carbides occurs This phenomenon takes placewhen upon annealing the martensite starts to release car-bon leading to the formation of carbides For an SS withcarbon contents above 008 wt pct precipitation of carbidesis expected to occur in less than 1 minute for temperaturesin the range of 700 degC to 900 degC[21] This means that thereverted austenite phase will be depleted in carbon leading

Fig 18mdashSize of second-phase particles as a function of annealingtemperature

Fig 19mdashEDS spectrum obtained from (a) the austenite matrix and (b) thesecond-phase particles The second-phase particles exhibit a significantincrease in carbon chromium and molybdenum with respect to the matrix




to a destabilization of the austenite phase A simple calcu-lation of the MS temperature reveals that for a case wherethe austenite matrix is without carbon the MS is raised to43 degC In addition as the AISI 301 SS used in this work hassome amounts of nitrogen it is also possible for the mar-tensite to reject nitrogen during annealing Taking into con-sideration the effect of nitrogen on the MS an additionalcalculation reveals that the MS temperature can be raisedto 150 degC Thus for samples annealed above 750 degC which

undergo pronounced carbide precipitation thermally inducedmartensite is likely to form on cooling

B Influence of Annealing Temperature onAustenite Grain Size

For annealing temperatures from 650 degC up to andincluding 750 degC the reverted austenite grain size increaseswith temperature From 750 degC to 800 degC the revertedaustenite grain size remains approximately the same How-ever the 750 degC sample has a larger deviation from themean than the 800 degC sample For the samples annealedat 850 degC and 900 degC a significant increase in grain size is

again observedThe results obtained at temperatures up to 750 degC and

above 800 degC show that the final grain size is dictated bya nucleation and growth process However at 800 degC thefinal grain size seems to be achieved by a different routeAn explanation for this difference in behavior can be foundin the two possible martensite-to-austenite reversion mech-anisms that affect the final microstructure and grain sizeie a diffusional reversion mechanism vs a shear reversionmechanism[256202223] The first is expected to producelarge deviations in grain size except for very long anneal-ing times The second which occurs within a narrow rangeof temperatures becomes a diffusional-dependent processonly after the initial shear reversion of martensite to high-dislocation-density austenite occurs

The microstructure of the 800 degC sample is consistentwith the shear-type reversion mechanism Previous work on austenitic and Fe-Si-Cr-Ni steels confirms that a shearmechanism can become active at annealing temperaturesclose to 820 degC[202223] As previously explained the firststep in the shear reversion mechanism is the transformationof martensite into austenite where austenite maintains thesame morphology of the martensite (ie a microstructurewith a high density of defects) Further annealing causesdefect-free subgrains and dislocation cells of austenite toform Eventually the subgrains grow and coalesce resultingin a recrystallized-type structure As shown in Figures 12

and 13 the reverted austenite grains have a higher densityof defects compared to all the other samples and evidenceof austenite subgrains is shown in Figure 13 As the 800 degCsample still shows a microstructure dense in defects and assubgrains are visible in the microstructure it can be arguedthat the subgrains in the 800 degC sample have not yet startedto grow

The similarity in austenite grain size between the 750 degCand 800 degC samples is not fully understood In the 750 degCsample smaller grains could be a consequence of carbideformation hindering grain growth In the case of the 800 degCsample the active mechanism is shear leading to a slowernucleation and growth of recrystallized-type austenite grains

In contrast to the sample annealed at 800 degC the speci-mens annealed in the range 600 degC to 750 degC appear to

Fig 20mdash(a) TEM micrograph of a M23C6carbide (b) Electron diffraction pattern of the carbide shown in (a) within the fcc matrixB5 frac12112

M23C6kB5 frac12011g

Table II Rockwell B Hardness Average Values for theAs-Cold-Rolled and All Annealed Samples obtained After

Performing Ten Measurements in Each Sample

As-coldRolled 600 degC 650 degC 700 degC 750 degC 800 degC 850 degC 900 degC

1195 1105 110 102 815 907 769 79




have followed a diffusional-type reversion mechanism Atthe lowest temperatures 600 degC and 650 degC the austenitegrains are still in the initial phase of the reversion which is

confirmed by the presence of martensite and small grainsizes with minimal deviation from the mean (Figures 5 7and 8)

For the sample annealed at 700 degC the reversion hasessentially been completed and has entered into the graingrowth stage Figure 9 clearly shows austenite grains thathave experienced significant grain growth (area denoted byL) and smaller austenite grains at the early stage of graingrowth (region denoted by S) For the 700 degC sample theaverage grain diameter increased to 083 mm and a moresignificant deviation from the mean (Figure 5) can befound This is consistent with the fact that the revertedaustenite has entered into the grain growth stage of the

reversionFor the sample annealed at 750 degC a significant rever-

sion of martensite to austenite has occurred The austenitegrains have completed the nucleation process and are wellinto the grain growth stage as exhibited by the averagegrain diameter of 174 mm which agrees well with theresults published by Di Schino et al[4] In addition the750 degC sample exhibits the highest ratio between the largestand the smallest grain sizes measured among all samplesThis is an indication of fast growth occurring in somegrains whereas other grains have just concluded the con-version process

At annealing temperatures of 850 degC and above a sig-nificant increase in grain size is observed As the shearreversion acts over a narrow temperature range the 850 degCsample may have formed austenite via a shear or diffusionalmechanism Since the annealing time used in this experi-ment was 30 minutes it is inconclusive as to whether at850 degC the subgrains formed as a result of the shear mech-anism grew and formed the recrystallized structure or thenucleated austenite grains resulting from a diffusional mech-anism followed the process of grain growth According toTomimura et al[5] the recrystallized-type austenite grainsize resulting from a shear mechanism was determined tobe larger than the recrystallized austenite grain size result-ing from a diffusional mechanism which would explain thelarger-than-expected grain size in the 850 degC sample As

both processes involve grain growth the deviation from themean does not offer additional clarification

The 900 degC sample was similar to the 850 degC sample interms of average grain size This similarity may be a resultof carbide precipitation limiting grain growth or a result of the 850 degC sample reverting via a shear mechanism andexhibiting increased growth of the subgrains The measuredaustenite grain size at 900 degC increased slightly to 58 mmwith a significant deviation from the mean (Figure 5) Theaustenite grains were equiaxed in shape with defects in

the form of annealing twins At this temperature the diffu-sional reversion mechanism is once again expected to beactive[202223]

In summary increasing the annealing temperature for agiven dwelling time increases the final reverted austenitegrain size The exception is the sample annealed at 800 degCwhich most likely underwent a shear reversion and has yetto complete the recrystallization phase as shown by thesmall subgrains and high number of defects (Figure 13(a))

C Influence of Annealing Temperature on Microstructure

Before discussing the effect of annealing temperature onthe overall microstructure it is important to evaluate themicrostructure of the CR specimen As previously dis-cussed the CR sample shows a microstructure composedby two distinct martensite structures regions of lath mar-tensite and regions of dislocation cell-type martensite Of the two types the predominant is the latter

Upon annealing there is no trace of retained dislocationcell-type martensite and only lath-type martensite is found(Figures 9 10 14 16 17 18 20 and 21) This result is anindication of fast conversion to austenite provided by thedislocation cell-type martensite This is not surprisingbecause heavy deformation of the lath martensite enlarges

the number of nucleation sites through the intrusion of slipbands and an increase in dislocation density

Thus for the samples annealed at 600 degC and 650 degC for30 minutes the remaining martensite regions are of thelath-type character (Figure 8(c)) In these areas the nucle-ation of austenite is slow At 700 degC more equiaxed aus-tenite grains (Figure 9) and a broader grain size variation(Figure 5) are apparent This broader variation is a result of austenite nucleation from the dislocation-type martensitestructure which starts earlier and leads to larger grainsand from the lath-type martensite which starts much laterand leads to smaller grains One other factor that can lead

to asymmetries in austenite grain growth is associated withthe severe carbide precipitation observed at 700 degC whichmay act as a barrier to recovery and grain growth This isparticularly true if carbide precipitation is fast enough suchthat carbides form on lath boundaries before the disappear-ance of lath martensite

Similar to the samples annealed at 600 degC and 650 degCretained regions of lath martensite still exist at 700 degC(Figure 10) which confirms the slow austenite reversionin these areas However as shown by the traces of slipbands in Figure 10 regions of martensite that have beenmore heavily deformed led to finer austenite grain size (leftside of white arrow in Figure 10)

At 750 degC it seems evident that the activation energy foraustenite nucleation is low enough to transform practically

Fig 21mdashVickers hardness and yield strength values for the CR and allannealed samples




all the martensite into reverted austenite (Figures 4 11(a))although some pockets of martensite remain (Figure 11(b))However due to the late transformation of the lath marten-site there is a significant variation in grain sizes (Figure 5)Austenite grains which readily nucleated from the disloca-tion cell-type martensite have been free to grow for quitesome time whereas austenite grains nucleated from thelath-type martensite have barely started their growth At750 degC the presence of carbides that have nucleated atthe lath boundaries can be seen and identified as strings

of particles (Figure 11)The sample annealed at 800 degC does not fit the general

trend discussed above Despite an increase in annealingtemperature of 50 degC the grain size is similar to the sampleannealed at 750 degC Furthermore it exhibits a much higherdensity of defects and fewer carbides (Figure 12 and 13)although the number of carbides in this sample is difficultto assess because the high density of defects may obstructtheir observation Nevertheless at 800 degC the mechanismof austenite reversion seems to be a shear-type mechanismAs shown in Figures 12 and 13 the microstructure exhibitsdislocation cells and subgrains typical of a martensitic

shear transformation Within this mechanism carbide pre-cipitation might be hindered because the martensite phasedoes not have time to release carbon prior to the formationof austenite

For the samples annealed at temperatures higher than800 degC the microstructure consists of recrystallized grainsof austenite decorated with carbides (Figures 14 and 16)The small increase in grain size between 850 degC and 900 degCseems to indicate that the presence of random carbideswhich grow and coarsen during annealing plays an impor-tant role in suppressing grain growth

One other important aspect shown in Figures 15 and 17 isthe presence of lath-type martensite at 850 degC and 900 degC

We argue that this phenomenon is caused by the partialtransformation of austenite back to martensite during cool-ing as a result of carbide precipitation and consequentdepletion of carbon from the matrix which results in a lessthermodynamically stable austenite

With respect to the precipitation of carbides the majoritywere found in the austenite matrix and not at the austenitegrain boundaries (Figures 11 14 and 16) However someof these carbides that appear to be within the grains arelikely to be carbides that nucleated at the lathndashmartensiteboundaries as these carbides commonly form string-likeformations (Figure 11) Nevertheless the massive presence

of carbides within the austenite grains is different from theexpected grain boundary M23C6 carbide precipitation thatoccurs in austenitic SS[2425] This may simply be a conse-quence of the fact that (1) austenite grain boundaries wereable to sweep past stationary carbide particles which pre-viously nucleated at interlath interfaces but have sincebeen engulfed by the moving austenite grain boundariesor (2) the presence of carbides within grains is predomi-nantly a result of the large number of defects such asdislocations subboundaries and twin structures introducedduring heavy deformation prior to the annealing temper-ature The latter seems more likely These defects act asnucleation sites for carbides during annealing which aredetrimental to the final microstructure as they deplete theaustenite stabilizing elements carbon and nitrogen promot-

ing the formation of thermally induced martensite uponcooling On the other hand they might be beneficial forsuppressing grain growth

D Influence of Annealing Temperature on Mechanical Properties

As shown in Table II and Figure 21 the samplesannealed at 600 degC and 650 degC show significantly highhardnessyield strength values This is due to the high vol-

ume fraction of martensite still present and the exhibitedsmall austenite grain size However a comparison betweenthe 600 degC and 650 degC samples shows that despite thesignificant increase in austenite volume fraction for the650 degC sample (Figure 4) the difference in hardnessyieldstrength is relatively small This seems to indicate that interms of hardnessyield strength the significant reduction inthe volume fraction of martensite for the 650 degC sample iscompensated by the small grain size exhibited by thereverted austenite For samples annealed at 700 degC and750 degC a drastic reduction in hardnessyield strength isnoted (Table II and Figure 21) As expected this is the

result of an increase in austenite volume fraction and par-ticularly an increase in austenite grain size (Figure 5) Thistrend is reversed for annealing treatments above 750 degC Asshown in Table II and Figure 21 the sample annealed at800 degC exhibits higher hardnessyield strength values thanthe 750 degC sample Two reasons for this behavior can beidentified (1) an increase in the volume fraction of marten-site for the 800 degC sample (Figure 4) due to the formationof carbides and consequent increase in the Ms temperatureand (2) a microstructure with a high density of defectstypical of a shear reverse transformation For the samplesannealed at 850 degC and 900 degC the hardnessyield strengthvalues are decreased with respect to the 800 degC sample

(Table II and Figure 21) This tendency is based on thedramatic increase in austenite grain size for the 850 degCand 900 degC samples (Figure 5) despite the larger volumefraction of martensite The significant effect of the auste-nitic grain size on the hardnessyield strength values isevident when comparing the 850 degC and the 900 degC sam-ples In this comparison although the martensite contentincreases with temperature the hardnessyield strength val-ues decrease due to an increase in grain size

Finally the yield strength values of all annealed samplesare compared with that of conventionally annealed AISI301 SS (dashed line in Figure 21) It is clear that all

annealed samples studied in this work exhibit yieldstrengths greater than those shown by conventionally trea-ted 301 SS The results are particularly interesting for the750 degC sample which exhibits essentially an austenitestructure but possesses at least twice the strength of conven-tionally treated 301 SS

Ultimately the goal of the work is to enable the industrialapplication of laser annealing and high-speed deformationto nanosubmicron-grained austenitic SS The work pre-sented here provides the basis for this innovative futurework While providing insight into the mechanisms involvedin producing nanosubmicron grains in an austenitic SSthere is still much to learn about the role of compositionthe amount of cold rolling and annealing time in the rever-sion mechanisms and texture achieved after annealing




V CONCLUSIONS

Based on the results obtained in this work it is possibleto conclude the following

1 A 90 pct cold reduction induces the formation of twotypes of martensite in AISI 301 austenitic SS lath-typemartensite and dislocation cell martensite Of the twothe predominant type is the dislocation cell martensite

2 Upon annealing AISI 301 SS undergoes a diffusionalreversion mechanism from martensite to austenite inall annealed samples except for the sample annealed at800 degC and possibly the 850 degC sample for which ashear reversion mechanism seems to be active

3 The highest reverted austenite content was achieved atan annealing temperature of 750 degC for a dwelling timeof 30 minutes as shown by the SQUID data At anneal-ing temperatures below 750 degC the microstructure con-tained retained lath-type martensite whereas at annealingtemperatures above 800 degC thermally induced marten-site was formed on cooling

4 The samples annealed at 600 degC and 650 degC exhibit thesmallest reverted austenite grain size of approximately

0250 mm and the narrowest grain size variation Themicrostructure is a combination of austenite and marten-site On the other hand the highest ratio between thelargest and the smallest austenite grain size is observedat 700 degC and is attributed to the interplay between theaustenite conversion and grain growth associated withthe diffusional-type reversion This result indicates thatthe completion of the transformation from martensite toaustenite occurs in less than 30 minutes and that asmaller austenite grain size can be achieved usingshorter annealing times

5 Above 800 degC there is a significant increase in grain

size (from about 1 mm at 800 degC to nearly 6 mm at 900 degC)and a decrease in the ratio between the largest and thesmallest austenite grain size The decreased ratio indi-cates that while the austenite grains are getting largerpractically all grains are now in the grain growth regime

6 TEM and EDS analyses show the presence of second-phase particles which are identified as (Fe CrMo)M23C6 carbides Regardless of annealing tempera-ture carbides are all of globular shape and found mainlywithin the matrix of reverted austenite grains Thematrix precipitation is attributed to a high density of defects present in the structure acting as nucleation sitesfor the carbides

7 Samples annealed at 600 degC and 650 degC exhibit hardnessvalues that are 70 pct of the CR samples Above 650 degCand down to 750 degC there is a sharp decrease in hardnessdue to an increase in austenite grain size and volumefraction of austenite At 800 degC the hardness increasesagain due to higher martensite formation and density of defects However above 800 degC the change in hardnessis predominantly dictated by the austenite grain size

8 To create austenitic SS with a nanosubmicron grainstructure an annealing temperature below 700 degC shouldbe used for a dwelling time of 30 minutes Howeverunder these conditions the microstructure consists of amixture of austenite and martensite which may haveconsequences for formability In this context futurework should concentrate on short annealing times fortemperatures above 700 degC

ACKNOWLEDGMENTS

The authors thank Mr Shreyas Rajasekhara for perform-ing the hardness measurements The authors at the Univer-sity of Texas at Austin would like to acknowledge thefinancial support from the National Science Foundation(NSF) USA Award DMR-0355234

REFERENCES

1 A Di Schino I Salvatori and JM Kenny J Mater Sci 2002 vol37 pp 4561-65

2 K Tomimura S Takaki S Tanimoto and Y Tokunaga ISIJ Int1991 vol 31 pp 721-27

3 N Tsuji R Ueji Y Minamino and Y Saito Scripta Mater 2002vol 46 pp 305-10

4 A Di Schino M Barteri and JM Kenny J Mater Sci Lett 2002vol 21 pp 751-53

5 K Tomimura S Takaki and Y Tokunaga ISIJ Int 1991 vol 31 pp1431-37

6 S Takaki K Tomimura and S Ueda ISIJ Int 1994 vol 34 pp 522-277 Y Murata S Ohashi and Y Uematsu ISIJ Int 1993 vol 33 pp

711-208 M Foldeaki H Ledbetter and P Uggowitzer J Magnetism Mag

Mat 1992 vol 110 pp 185-969 Research Laboratories Outokumpu Oy Finland unpublished research

200410 AH Eichelman Jr and FC Hull Trans ASM 1953 vol 45 pp

77-10411 T Angel J Iron Steel Inst 1954 vol 177 pp 165-7412 M Stalder S Vogel MA Bourke JG Maldanado DJ Thoma and

VW Yuan Mater Sci Eng A 2000 vol 280 pp 270-8113 MJ Dickson J Appl Crystallogr 1969 vol 2 pp 176-8014 SSM Tavares PDS Pedrosa JR Teodosio MR da Silva JM

Neto and S Pairis J Alloys Compd 2003 vol 351 pp 283-8815 D Tabor Hardness of Metals Oxford University Press Oxford

UK 95116 M Tokizane N Matsumura K Tzuzaki T Maki and I Tamura

Metall Trans A 1982 vol 13 p 137917 H Araki K Hirata and Z Fujimura J Jpn Inst Met 1980 vol 44

p 124418 R Ueji N Tsuji Y Minamino and Y Koizumi Acta Mater 2002

vol 50 pp 4177-8919 P Marshal Austenitic Stainless Steels Microstructure and Mechanical

Properties Elsevier Applied Science Publishers New York NY 198420 YK Lee CH Shi DS Leem JY Choi W Jin and CS Choi

Mater Sci Technol 2003 vol 19 pp 393-9821 KC Russell and HI Aaronson Precipitation Process in Solids Met-

allurgical Society of AIME Warrendale PA 197822 K Tomimura S Kawauchi S Takaki and Y Tokunaga Tetsu to

Hagane 1991 vol 77 p 151923 K Tomimura S Takaki and Y Tokunaga Tetsu to Hagane 1988 vol

74 p 164924 R Stickler and A Vincknier Trans ASM 1961 vol 54 pp 362-8025 R Beltrain JG Maldanado LE Murr and WW Fisher Acta

Metall 1997 vol 45 pp 4351-60



steels have shown correlations between amount of coldwork annealing temperature and dwelling time on themicrostructure achieved very limited work has been doneon commercial austenitic SS alloys containing carbon andnitrogen[127] In addition most of the research performedso far has dealt with the effects of annealing on the rever-sion mechanisms but a strong correlation between the grainsize and the microstructure achieved has not been investi-gated In particular a fundamental understanding of theeffects of the annealing temperature on the microstructuremartensite-to-austenite (a9 g ) ratio austenite reversion andaverage grain size has not yet been accomplished

The goal of this paper is thus to correlate the annealingtemperature with the reverted austenite grain size phasefraction and microstructure achieved in a commercial AISI301 SS via magnetic measurements X-ray diffractometry

and transmission electron microscopy (TEM) experimentsA commercial AISI 301 metastable SS grade was selectedbecause its metastable austenitic structure facilitates theformation of stress-induced martensite and subsequentlypromotes an efficient austenite reversion In addition AISI301 SS grades are commercially available thereby widen-ing the potential impact of this research work

II EXPERIMENTAL PROCEDURE

A Materials

The AISI 301 SS used in this work was provided byOutokumpu Stainless Oy Finland with an alloy composi-tion (wt pct) given in Table I This SS has an Md30 tem-

perature of 18 degC[89] and a calculated MS temperature of ndash118 degC[1011] The material was produced by continuouscasting and subjected to hot rolling heat treatment and afinal 90 pct cold-rolling reduction at Outokumpu StainlessOy The amount of retained d-ferrite was found to be appro-ximately 02 vol pct As-received 90 pct cold-rolled sheetswere cut into rectangular specimens subsequently subjectedto a heating rate of 100 degCmin and annealed isothermallyfor 30 minutes at the temperatures 600 degC 650 degC 700 degC750 degC 800 degC 850 degC and 900 degC followed by forcedair-cooling For comparison one specimen was left in theas-cold-rolled (CR) state

B Methods

1 X-ray diffractionRectangular sections approximately 5 3 7 3 08 mm3

were cut with the rolling direction (RD) parallel to the longdimension These samples were subsequently measured in aPhillips PW1720 X-Ray Diffractometer (XRD) operating ata voltage of 40 keV and a current of 40 mA The tests wereperformed at ambient temperature using CuK a1 radiationPeak measurements were taken from 30 to 90 deg (2u) insteps of 01 deg with a dwell time of 4 seconds From theresulting intensity vs 2u plots MDI Jade 65 software was

used to index the peaks The results were compared topublished XRD data for an austenitic SS[12]

In addition the X-ray spectrum of the CR specimen wasused to calculate the volume fraction of austenite and mar-tensite in this sample based on the method published byDickson[13] According to this method the volume fractionsof austenite Cg and martensite Ca were evaluated on thebasis of the first three reflexions for the austenite ((111)(200) (220)) and the martensite phase ((110) (200) (211))in the form[13]

Cg

Ca

5

1

ng

+ ng

0

Ig

Rg

1

na+na

0

Ia

Ra

[1]

where Ca frac141= 11ethCg =CaTHORN

Cg frac14ethCg =CaTHORN= 11ethCg =Ca

na and ng are the number of martensite and austenite reflex-ions considered Ia and Ig are the integrated intensities of martensite and austenite for the reflexions considered andRa and Rg are the relative intensities of martensite andaustenite for the reflexions considered On the basis of Eq [1] the volume fraction of martensite present in theCR (90 pct reduction in thickness) sample can be calculatedas 974 pct This sample was selected as the reference valuefor the determination of phase fraction in all the samples by

Fig 1mdashNovel method used to produce nanosubmicron-grained austenitewith good formability properties

Table I Chemical Composition of AISI 301 Stainless Steel(Weight Percent)

C Cr Ni N Mn Si Mo

0096 167 66 00635 123 118 072

P Cu Co S D Y Fe

026 017 011 0001 00031 0005 Balance



















III RESULTS
















C TEM Observations



























































circle B5

frac12111a0kB5

frac12011g





























IV DISCUSSION









than 118 deg


[256181920]



















M23C6kB5 frac12011g




1195 1105 110 102 815 907 769 79















































V CONCLUSIONS












ACKNOWLEDGMENTS


REFERENCES






















Metall 1997 vol 45 pp 4351-60


















III RESULTS
















C TEM Observations



























































circle B5

frac12111a0kB5

frac12011g





























IV DISCUSSION









than 118 deg


[256181920]



















M23C6kB5 frac12011g




1195 1105 110 102 815 907 769 79















































V CONCLUSIONS












ACKNOWLEDGMENTS


REFERENCES






















Metall 1997 vol 45 pp 4351-60










C TEM Observations



























































circle B5

frac12111a0kB5

frac12011g





























IV DISCUSSION









than 118 deg


[256181920]



















M23C6kB5 frac12011g




1195 1105 110 102 815 907 769 79















































V CONCLUSIONS












ACKNOWLEDGMENTS


REFERENCES






















Metall 1997 vol 45 pp 4351-60
















































circle B5

frac12111a0kB5

frac12011g





























IV DISCUSSION









than 118 deg


[256181920]



















M23C6kB5 frac12011g




1195 1105 110 102 815 907 769 79















































V CONCLUSIONS












ACKNOWLEDGMENTS


REFERENCES






















Metall 1997 vol 45 pp 4351-60




























IV DISCUSSION









than 118 deg


[256181920]



















M23C6kB5 frac12011g




1195 1105 110 102 815 907 769 79















































V CONCLUSIONS












ACKNOWLEDGMENTS


REFERENCES






















Metall 1997 vol 45 pp 4351-60














M23C6kB5 frac12011g




1195 1105 110 102 815 907 769 79















































V CONCLUSIONS












ACKNOWLEDGMENTS


REFERENCES






















Metall 1997 vol 45 pp 4351-60














































V CONCLUSIONS












ACKNOWLEDGMENTS


REFERENCES






















Metall 1997 vol 45 pp 4351-60

influence of annealing treatment on the formation

Documents