FID-RI33 428 A METALLOGRAPHIC INTERPRETATION OF THE OLT HEAT /TREATMENT OF FERRITIC CRY..(U) CALIFORNIA UNIV BERKELEYDEPT OF MATERIALS SCIENCE AND MINERA. J I KIM ET AL.

UNCLASSIFIED MAY 2 TR-3 N614 75 C 8i54 F/6 111'6

1111.0 LI I 6

Ji

L

, MICROCOPY RESOLUTION TEST CHART~NATIONAL BUREAU OF STANDARDS-1963,-A

| I ,, mflJI

A METALLOGRAPHIC INTERPRETATION~ OF THE QLT HEAT TREATMENT

OF FERRITIC CRYOGENIC STEELS

"rki

J. 1. Kim and J.W. Morris, Jr.

Department of Materials Scienceand Mineral Engineering

UNIVERSITY OF CALIFORNIA, BERKELEbTMAY, 1983 j EEC-

-- Technical Report No. 13Of fice of Naval Research

Contract No. N00014-75 -C -0154This docurr,t has bee1 n aIpprovgd

for Pubfi- rcTa-?e (Ind sce qt

4'4L

icFILE COPY 83 09 30 018

SECUP:TV CLASSIFICATION OF TNIS PAGE (Whon Dom Entered)

REPORT DOCUMENTATION PAGE READ INSTRUCTIONS""_ _ __BEFORE COMPLETING FORMI REPORT NUMSER 2. GOVT ACCESSION NO. 3- RECIPfENT'S CATALOG NUMBER

#134. TITLE (and Subtitle) S. TYPE OF REPORT & PERICO COVERED

A METALLOGRAPHIC INTERPRETATION OF THE QLT 6 PERFORMING ORG. REPORT %UMBER

HEAT TREATMENT OF FER.ITIC CRYOGENIC STEELS7. AuTb4qOmR) 6 CONTRACT OR GRANT NUMIER(a)

J. I. Kim and J. W. Morris, Jr. N00014-75-C-0154

9 PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM EL-MENT. PROJECT, TASK_.9, ARE& & WORK UNIT 04UMSECRS

University of California, Berkeley AU

Department of Materials Science & Mineral Eng. NR 031-762Berkeley, California 94720

11. CONTROLLING OFFICE NAME AND ADDRESS IZ. REPORT DATE

NAVAL SYSTEMS COMMAND may, 1982Washington, D.C. 20362 .NUMIEROF PAGES5

14 MONITORING AGENCY NAME G AOORESSOtI differet! from Controllind Offlce) IS. SECURITY CLASS. (01 tle Popoff)

OFFICE OF NAVAL RESEARCH Unclassified800 N. Quincy StreetArlington, VA 22217 sOECLASSIICATION, OWNGRADINGSCMEOULE

16. DISTRIBUTION STATEMENT (at thl Report)

17. DISTRIBUTION STATEMENT (of the abstract entered In Slock 20, It dllr- grom Report)

18. SUPPLEMENTARY NOTES

Presented at the International Cryogenic Materials Conference, Kobe, JAPANPublished in "ICMC, Kobe, Japan, 11-14 May 1982", Butterworth, England, 1982.

IS. KEY WORDS (Continue on revaree aide II neceseary and Identify by block number)

Steel, Fe-Ni Alloys, Cryogenic Structural Materials, Heat Treatment,

Fracture, Ductile-Brittle Transition

20. ABSTRACT (Continue on reverse aide it neceesry and I'entfy by block number)-%;Commercial 5-6NI steels are toughened for service by a 3-step heat treatment

which is sometimes designated the .'QLT" treatment. The present investigation

was undertaken to determine why this treatment is necessary and successful. It:was concluseded that the multi-step heat treatment is necessary because of thelow nickel content. The intercritical anneal (L) serves to create regions of i.high solute contents along the martensite lath boundaries. The intercriticaltemper (T) then precipitates a dense distribution of high solute, stableaustenite within these enriched regions..

DD I 1473 tolON or I NOv 6s is OSO,ETES, N 0102 LF 014 6601[; *1 ;. ," • * • , -a- .• . . . . ... . . . . .. . . .flr¢l: l['¢ CLA JI. F I CA TI 1 0 .OR; ,%

,- - a . . -. s . . - 4, -•

A ErAoAIEIC D or 722 N.SM 7EAT OF E MI1TIC CZOMIIC flUB

by

3. I. Kim* and 3. W. Morris, Jr.

Department of Materials Science and Mineral EngineeringUNIVESITY OF CALIFORNIA, DKE

- A080s0ion For

UNVi TS' G-RA&DI,C T ASU!'announced 0

Justifisation.-

_Distribution/1va~labi'ty eodes

!Aveil and/orVDist Speatal

Q, 4

44

C Now at the Thomas . Watson Research Center, IBM, Yorktown Heights. New York

, W. , :-. r .. .. , ,' . ".'.'. -... .-.. .... . . -.. . - - .- ." -* -_- . -.. ' -., . . -.......... o......... .... ..... ... ........... >.. ..... -. .....-'.5, " , . " - . ", -*. - • .* • . . . . * • . . . . 5. .- -* ..

__ . .. - io*. .-- ,, -_%. . o. -. . .- ..- -. _ - . . .-. '."i .- , •'.. .. . " i'

2

INTRODUCTION

St

Ferritic steels which are intended for structural use at cryogenic tempe-ratures are thermally processed to lower the ductile-brittle transition tempe-rature to below the intended temperature of service. In conventional 9Nisteel this treatment involves a straightforward intercritical temper at arelatively low temperature within the two phase (a + y) region. During theintercritical temper m fine, dense distribution of the austenite phase isprecipitated along the boundaries of the dislocated laths of the martensite

* structure. This austenite is thermally stable on cooling to at least 77K. Itlowers the ductile-brittle transition temperature by breaking up the crystal-lographic alignment of the martensite laths within packets [1]. Ferriticsteels of lower nickel content can also be toughened for cryogenic service(2,31 as can ferritic iron manganese steels containing no nickel at all (4,5].But in these cases more elaborate thermal treatments are required. The mostcommercially important of these alternative heat treatments are required. Themost commercially important of these alternative heat treatments is the QLTtreatment, which is a 3-step heat treatment employed for toughening the 5-6Nisteel developed some years ago by the Nippon Steel Corp. in Japan and by ARMCOsteel in the United States. The QLT treatment is the fourth of thosediagrammed in Fig. 1. It involves two sequential intercritical heat treat-ments. The first of these (1) is an intercritical anneal in the upper rangeof the two phase (a + y) field. The second (T) is an intercritical temper atlower temperature within the two phase field.

While the QLT treatment has been used with commercial success for someyears, the fundamental reasons for its effectiveness remain unclear. Toclarify the microstructural sources of toughening we have studied the inter-play of composition and microstructure in determining the ductile-brittletransition temperature benefit achieved from the QLT treatment. The alloyused in this investigation was a commercial grade 5.5Ni steel which was pro-vided by the Nippon Steel Corp. Its composition was determined to be: Fe-$.86N-1.21N-0.69Cr-0.20o-0.2Si-0.06C-0.01S-o.08P. The alloy was annealedat 12000 C for two hours to remove the effect of prior thermal mechanicaltreatment. It was then solution annealed at 9000 C for 2 hours. Experimentalsamples were out from the annealed plates and subjected to one of the fourhost treatments diagrammed in Fig. 1. The microstructure of the alloys wasstudied as a function of heat treatment using optical, transmission and scan-ning transmission electron microscopy. The latter technique was used todetermine the chemical composition of the austenite present in the heat-treated samples [6]. The volume fraction of precipitated austenite was deter-mined by x-ray diffraction as suggested by Killer (7] and by backscatterMossbaser spectroscopy using the apparatus and techniques described by Fultz

RESULTS

The volume fraction of austenite in the microstructure at room tempera-ture and after chilling to 77K is tabulated as a function of heat treatment inTable 1, together with the concentration of detected solutes in the austenito,

- r, % _'* -- ,' , , % , .%-. - . . -*% " ". .• -. - .- - - . -. . -. . • . • -. -. -. -. - - -. • . -. . "

S" " " -" "- "" "" " " ' " '-"" . . . .

~'% ..................

3

as determined by STEM measurements [6]. The variation of microstructure withheat treatment is described in detail in Reference 9. The results may besummarized as follows:

1. The Q condition. In the as-quenched condition the alloy has a dislo-cated lath substructure in which martensite laths are organized into alignedpackets which subdivide the prior austenite grains. The martensite lathswithin a given packet are almost always in a single variant separated by low-angle lath boundaries.

2. The Microstructure. In the QT condition the primary constituent 7-:is martensite whose lath boundaries are decorated by a mixture of retainedaustenite and fresh martensite (from retransformation of the austenite nuc-leated during the tempering step). The austenito is retained in relativelylow volume fraction and has the Kurdjemov-Sachs relation to the parent marten-site. This austenite is relatively lean in substitutional solute content.Its low solute content and small volume fraction are apparently due to acombination of the low tempering temperature, the relatively short temperingtime, and the low diffusivity of the substitutional species which preferenti-ally segregate to the austenite phase. The low solute content is presumablyresponsible for the thermal instability of this austenitep. the bulk of itreverts to martensite on cooling to 77K• .

3. oTh Microstructure• A 100 hrs heat treatment gives the struc-ture shown in"Fis. 2. The tempered martensite retains a fine polygonizedsubstructure. The austenite is present in much higher volume fraction, andthe precipitated austenite particles have coarsened and spheriodized into moreequiaxed subgrains of approximately 0.5p diameter. This austenite is muchricher in substitutional solute content (Table 1) and is thermally stable. Itis almost entirely retained after the sample has been cooled in liquid nitro-Son.

4. The OL Kicrostructure. Increasing the heat treatment temperature to6700C for 1 hour yields a microstructure containing three distinct microstruc-tural constituents. These are: 1) tempered martensite having a fine, poly-Sonized structure, 2) fresh, highly dislocated martensite which presumablyarises from the transformation of the greater part of the austenite precipi-tated during the intercritical anneal and 3) retained austenite. The auste-site is not as rich in substitutional solutes as that after QTI00 and must beexpected to be lower in carbon content because of its higher temperature offormation. It is thermally unstable with respect to martensitic transforma-tion on cooling to 77K (Table 1). Given the combination of extensive chemicalredistribution and cycling phase transformation during the L treatment onemight expect a substantial decomposition of the martensite packet structure.However, this is not the case. The cyclic c-y>a transformation has a strongmemory. When precipitated austenito reverts to martensite on cooling, italmost invariably returns to the martensite variant which Save it birth, withthe consequence that the martensite packet size remains virtually unchanged.

. . .

4"1

Table 1. Composition and Volume Fraction of Precipitated Austenite.

Heat Volume Fraction (5)Treatment Fe Ni Mn Cr Si 298K 771

89.5 6.6 2.2 1.2 0.1 5.9 2.8

ro100 85.0 9.1 3.8 1.5 0.2 8.2 7.9

QL 86.1 8.4 3.7 1.3 0.1 8.8 2.0

QT 82.8 8.8 4.3 3.2 0.5 9.2 8.8

Base Alloy bal. 5.86 1.21 0.69 0.2

S. The QLT Nicrostruotur. When the QL treatment is followed by a temperat 6000 C for 1 hour. the result is the QLT treatment. The microstructure ofthe steel after this treatment (Fig. 3) consists of three constituents. The Ltreatment causes a ohemical decomposition of the alloy into regions which arerelatively rich or relatively poor in solute content. The subsequent T treat-meat causes a secondary decomposition of the solute-rich islands. The finalmicrostructure consists of a low solute martensite matrix containing elongatedregions of solute-rich tempered martensite which themselves contain elongatedsubregions of high-solute precipitate austenite with a slight admixture offresh martensite. The austsnite phase is shown in the dark field micrographin FIX. 3. Its composition is very nearly equal to that of the austeniteformed after 100 hours tempering at 6000C. Virtually all of this austenite isretained on refrigeration at 77K.

The microstructural changes which occur during heat treatment have almostno effect on the tensile properties of the alloy, but they profoundly influ-eonce its impact toughness. The Charpy impact energy is plotted as a functionof temperature for each of the four heat treatments in Fig. 4. The evidentchanges include the pronounced increase in the upper shelf Charpy energy afterthe QL treatment, and the dramatic decrease in the ductile-brittle transitiontemperature in the QLT condition.

DISCUSSION AND CONCLUSION

Previous research [1,9,101 suggests that intercritical heat treatmentinfluences the toughness of ferritic cryogenic steels primarily because of theprecipitation and retention of the austenite phase. The precipitation ofaustenite can affect both the level of toughness above the ductile-brittletransition and the value of the ductile-brittle transition temperature. Thelevel of toughness above the ductile-brittle transition principally reflectsthe distribution of carbon and carbides, which can be effectively Betteredduring austenite precipitation. The carbon redistribution is the probable

V .? .~. . . . . . .

*~'~ ~9.. ~ ~ *~* . .* .. . .. *. . . .

cause of the dramatic increase in the upper-shelf toughness after the QLtreatment [10]. The decrease in the ductile-to-brittle transition temperaturereflects the role of austenito in zefining the effective grain size of thealloy by breaking up the crystallographic alignment of martensite laths in thepacket structure.

To accomplish grain refinement the precipitated austenite must have twoattributes: it must have sufficient thermal stability to be retained at thetesting temperature (though it ultimately transforms during mechanical defor-nation [1,91 (and it must be sufficiently dense in distribution to accomplishgrain refinement. The austenite introduced by the QT2 and QL heat treatmentshas insuffjicient thermal stability. The QT100 treatment introduces a ther-mally stable austenito phase, but the austenito is blocky and widely dispersedthrough the matrix. Such an austenite is ineffective in grain refinementsince it is relatively easy to find transpacket cleavage paths which avoid theaustenite entirely. The QLT treatment, on the other hand, establishes anaustenite phase which is both thermally stable and dense in its distributionbecause the precipitation of austenite in the T treatment occurs in interlathregions which have already been enriched in solute species by the prior Ltreatment.

REF ENCES

1. I.1. Morris, Jr., C.K. Syn, J.1. Kim and B. Fultz: Proc. of the Inter-nat'l Conf. on Martensitic Transformations, MIT Press, Cambridge, MA(1979), pp. 572-577.

2. S. Nagashima, T. Ooka, S. Sekino, H. Minura, T. Fujishima, S. Yano and

B. Sakurai: Trans. ISIS, vol 11 (1971), pp. 402-411.

3. S. Nagashima, T. Ooka, S. Sokino, H. Mimura, T. Fusishima, S. Yano and

. Sakurai: Tesu-to-Hagane, vol. 58 (1972) pp. 128-141.

4. K. Niikura and S.W. Morris. Jr.: Met. Trans. A, vol. 11A (1980) pp.1531-1540.

S. S.K. Hwang and 3.1. Morris, Jr.: Met. Trans. A., vol. 11A (1980, pp.1197-1206.

6. 3.I. Zim and S.W. Morris, Jr.: Met. Trans. A. vol. 12A (1981) pp. 1957-

1963.

7. R.L. Miller: Trans. ASM, vol. 57 (1964) pp. 892-899.

8. B. Fultz: M.S. thesis, Department of Materials Science and Mineral

Engineering, University of California, Berkeley (1978).

9. 3.1. Kim, C.K. Syn and S.W. Morris, Jr.: Met. Trans. A, vol. 14A (1983)pp. 93-103.

10. 1.1. Kim and 3.1. Morris, Jr.: Mot. Trans. A, vol. 11A (1980) pp. 1401-1406.

I hr

S 2 hrs

Qr2?Fe Ni .

.4 I 800C Ihr

,OO rOOI 60"t h00

67 'O , hr :_ _ _ . Ihr

I Ihr

. I hr_ -__ - _ rit

XBL802-4640FIGURE 1. The Four Heat Treatments

'i~~~~~~~~~~~.. %. . .................... . ............ . . "..... . "." .. " . ,• .. . . . .

'A-'C

" On.4

-sr

RX)m

i IlJl. (Illm

* (IT)

0 lll3 Martensite0 Eilo] Austenits

FIGURE 2. The QT 0 Microstructure

L 'i . ",-" , ." - ";'" . " ". . . . .. • ... . . . .

-r , ,,, W , ,,£ "<.. ¢ . .' .. . . *.' -'*.' .,.* *.., ' -.' ." *. . . " .' . . -' . . " " " . ' . ' . . Z .'

*~~ ~ .. - - . -. ' r -r r

FIGURE 3.The QLT Microstructure

S200

'~160-C.

Cw 120-

-0

-40

0 1

, ~ 40 -- • L

o I I-200 -150 -100 -50 0 50

Temperature (*C)

XBL 815-5664 AFIGURE 4. Ductile Brittle Transition Behavior

: :. .,,, ..... / ,;.'' .. .-. 9 ... . .,. ". .. ,..... .. .. :. . . . .. .