I

NASA TECHNICAL MEMORANDUM NASA TM-75824

ELECTRON_MICROSCOPE INVESTIGATION OF _e + y I;,_:

_m_FOmATIONSINANIRON-_c_m_mY I _3

V. A. Lobodyuk, L. G. Khandros, N. P. Fedas

Translation of ,Elektronnomikroskopicheskoyeissledovaniye e-_ Y preTrashcheniya v splavezhelezo_nikel," Metallofizika, No. 27, 1970p.i13-120.

NASA-TM-75824 19810013630 I_OR "_T_'_IY:'_"IZ"N_;'r":_

,,)Oev1.9/980

LA/yG/'_"f _td"S£ARCI'I CENTgRL'_SR/_RY,NASA

HAMILTON, VIRGINI _

NATIONAL AERONAUTICS AND SPACE ADMINISTRATION_ ._WASHINGTON, D.C, 20546 APRIL1980 -_-

/ '

STANDARD TITLE PAGE,. RU't'tf-to. 2. Go..,!'m.,,' Aee...lo" N•• ): R.elpl.n'·. C•••I., No.NASA TM-7S824

"5. R.po;' D•••~. Tltl••n4 $"b'ltl.

:Jn~ctron- /microscope ex + y \IIP'RTT lQAn

An study of 6. P.,fo,mln. O,••nl~.'I.n C04.I --transformations in an iron-nickel alloy(~...-..-

7.' "",ho,h) •• P.,'o,mln, O,.onl ••t1on R.port No•'-----------_._---_.- ~- '--- --._--.-----._.- -.----------_._---_ .... . ---_.-- - ---

\,.

V. A. Lob odyuk , L. G. Khandros, N. P. Fedas' 10. Wo,k'Unlt No..~, P.,fo,min. O'lloniaohOn Noma and Add,o.. 11. Con',oc' 0' G,.n' No.

.SCITRAN NASW-3l98Box·S4S6 13. T". o' Ropo" en. '0'104 COYO,."

Santa Barbara, CA.93l08 Translation '"

12. ,$pon.o,lnll A.oney Noma ond Add,...

National Aeronautics and Space Administration 1•• "n•.,I". A"oney Co4oWashington, D.C. 20546

U. $"pplomon'.,y No•••

Translatin ,0f'· "Elektronnomikroskopicheskoye issledovaniye~y prevr.ashcheniya splave .zhelezo-nikel, " MetallG- .'

Ct vfi'zika, No" 27, 1970, p. 113-120 9A70-36204)

-16, Abstroc, .. .

, . ~!~I"

,',',',.transformatiotlsThis is.a study to investigate a. + y , iri

an iron-nickel alloy, .Procedur'es used ape lJre s..~nted. Some resul1:!s are dispussed.. ,

Seven ~lectrori~miG~oScOpepictures .i'ncluded ...-,

"

" .' .. . ,., ..

r ....

", ..

\7. Koy Wo,d. (SeleCled by AUlhOr(s») 18. Oi."lb".lci" $'o'omon'

.UnlimitedUnclassified -

-' ..

?.'9. ~<. 20. Sec",lty CI.·..I•• C., 'hi. p.".) 21. N•• o' P •••• 22-Soc",lty Clonl'. (0' 'hi. ,op."), ..

12Unclassified Unclassified

'.~. '

.',

J,

.." '

ii

, Nf\SA-tIQ.4 ,.IM#4 J:;:O"':.$i.(lfNiI,h$ .4.lw:;t",;;;e.:;,4J.0%.;i¥¥¥_f¥@#Mt,!¥!4?¥444MN#¥MH,4I@@i£fW::,;;I'.W,1!J!.#"UH'('*'t4,R3$¥Y::SJJ¥FI""ljS"';<i"1i(;~?'.'

An Electron microscope_iStudy• of _ _ _ - Conversion in an

Iron-Nickel Alloy }

V. A. Lobodyuk, L. G. Khandros, N. P. Fedas.i 'L

The question concerning the kinetics and the mechanism /l___13*

of __ €onversi0n_in iron-nickel alloys has be4n repeatedly

investigated and discussed in the literature [i-4]. Recently

it has acquired considerable signiflcance due to the develop-

ment of ne_ alloys on his base and the application of harden-

ing treatment by means of direct and _e_erse 7_m_V conver-

sion,J

It is well Mnown _hat in a broad interval of concentrat-

ions _-_ conversionin these alloys_is due to the marten-

site mechanism. The formation of a contour onthe polished

surface of a sample containing an _-phase after heating

above the temperature of the beginning of _7 -conversion

isevidence of this. Moreover, the contour issimilar to that

which appears during cooling as the result of V_= conver-

sion [2]. The martenslte character of the reverse conversion

is also confirmed by i_s distribution over a broad temperat-

ure interval.

In addition, it blasbeen shown that along with the mar-

tenslte mechanism of m,*? conversion, a certain part of the

7_phase is converted into the s-phase by the_diffusion mechan-

ism. Thus in [5] it was established for a number of iron-nick-t

el alloys, contaln_ng from 9.5 to 18% Ni, that' m_ conver-

sion was accomplished by a quick acting martens_te mechanism

and then by a diffusion mechanism whose rate depends essen-

tially on the temperature and the composition. In the same

paper a similar conversion process is reported for an alloy

...... containing S0.9% Ni.

..... During conversion by the diffusion mechanism, diffusion

..... * Numbers in margins _indicate fore!gn pagination.1

of nickel from the _.phase Into the _phase occurs wlth sub-

sequent fusion of the supersaturated _solid solution into

two equlllbrlum _u an_ _-solid solutions (the first of which

Is poor In nickel, and the second, rich). After cooling to

room temperature, the_phase, which is rlch in nickel, was

converted Into the _rphase in accordance._with the metastable

diagram in the corresponding temperature region. When heated

to 450°C the %-phase _hich precipitated out was so rich in

nickel that it did not convert to the _phase at room tempera-

ture [5].

/114E

Using X-rays, we observed an increase In the amount of the

_phase during isothermal holding of an alloy wlth 32% Ni in

the 400-425eC temperature range. The samples were cooled in

advance ia liquid nitrogen. After heating briefly to 500°C noE,

traces of the _Tphase!lines could be detected on the X-rays.

To detect small amounts of the _phase which might still re-

main unconverted, the electron microscope method of examining

thin films for b_eaks was employed. An iron-nickel alloy wlth ...............

32% Ni concentration _nd initial martenslte conversion temp .............

erature of -60°C was selected for the study. The conclusion

of the reverse con,version for this alloy lles below 450°C.

Samples in the form of 0.5 mm thick sheets were quenched from

1000°C and cooled to -i96°C. Then they were placed in an oven

and kept there until completely warmed up (4 min) at different

temperatures, After being snb_.ected to heat treatment, the

samples were thinned down to 0.i mm by a grinder and polisher.

Disks of_diameter 3 mm were cut out of these sheets by the

electric spark method:_nd then by electropollshlng, thln foils

were obtained for scanning in the electron microscope UEMV-100.

On the electron d!iffraction patterns obtained from the

_. hardened samples only_ghosts of the face-centered cubic lat-

tices were observed. _After cooling to -196°C, martensite con-

2

i

version occurs, which was easily observed both on the X-rays,r

[as well as by means o_ the electron microscope. The _--phase

yield is nearly 80%. ,

I

In samples heated to 500°C we observed mainly the struc-

ture of the _-phase which contains a large number of disloca-

tlons. However, regions with an .mLphase structure were also

encountered comparatively often. Figure 1 contains a photo-

micrograph of such a region (x46000). On the photograph broad

twins in the _.phase (so-called deformatiOnal twins) and dis-

locations are visible, , In the upper right corner an electron

diffraction pattern is shown which was taken from this parti,

of the sample. All s_ots of the electron diffraction pattern

are contained in the body-centered cubic lattice. Spots be-

longlng to the face-centered lattice were not detected.I

Using the s_ng%e-suPface method of analysis, we deter-

mined the plane and di_ectlon of twlnmlng. It turned out that

defermatIQnal twinnlng occurs according to the system (12_1_

.ll_ll_ .

Besides the defermat!onal twins, we alse observed thin

twins (the so-called conversion twins), traces of which arel

on the electron diffraction patterns. In Fig. 2 (x57000) the

structure of a portion of a martensite crystal with thin twins

is shown, and in the upper left [?] corner is an electron

diffraction pattern of this portion. In it only ghosts of the

_phase a_e observed. _Besides the basic ghosts, the ghostsof the twins and traces are visible. The traces are located

along tTI0_ :d, perpeDdlcular to the direction of the thin

twins. Stereegraphic analysis has shown that thin twinning

occurs due to the system II_l_ lllll__ .

Sometimes a dislocational network was observed (Fig.

3, x40000) with two systems of almost parallel dislocations.

3¢i'

v

Figure 1 ..........

Figure 2

Figure 3 /116

i

r

Figure 4

_ 5

In the upper right corner is located a micro electron dlff- /117

faction pattern of thls pertlon. In it ghosts of the Q-phase

are visible. One system of dislocations lies in _ll_)_01|_ ,

and the other in (11f)_I1101_ •

?

Figure 5 Figure 6

After heating the alloy to 600@C the number and magnitude

of the portions with an _-phase structure were substantially

reduced. However, wewere able to obtain micro electron dif-

fraction patterns on Which onlyghosts belonging to the D-phase

were nevertheless visible. On Fig. 4 (x29000) is a photomicro- I

graph and the correso_ending micro diffraction pattern. On

the picture thin twins with high dislocation density are visi-

ble which have formed as the result of _o_ _ - conversion. As

the temperature increases further, the number of portions with

an _-phase structure decreases more and more; however, at 700

and 800°C, along with_o?--phasespots on certain micro electron

diffraction patterns, ghosts are detected whlc_ belong to the

body-centered cubic lattice.

On Figs. 5 and 6 there are photomicrographs taken of

samples after heating to 700 and 800°C, respectively.

Taking into account the fact that after heating the alloy

above 700°C we are unable to obtain electron photographs with

ghosts whlchbelong to the_-phase only, we can hence con- /118

clude that the _-phase regions at these tempe'r'aturesare small-

er than the dimensions of the portion from which the microdif-

fraction was taken (5-7 mc).

The presence of _. and _--phase ghosts on _he same electron

photograph makes it possible to determine their mutual crystal-

lographic orientation. In this case there are the following

orlentatlonal relations between the _- and ,Y--phases.

[loon li [_i7o1_,lOlOlv Ii 11101o,,

':] [O01/v l; IO01l_,.

Figure 7.

However, in certa_In cases it was ebserved that the direc- •

tlons with low indices" were not parallel. Thus in Fig. 7 [I00]_

makes an angle of 7° with JITOI_ . The reasons for this will

be stated below.

The interrelatlon_hip between the m_ and _ophase lattices

defined here la the tYPical mutual Bain's orientation [6]. This

orientation was observed in a Fe-30.2% Ni alloy under compres-

sion along one of the cubic axes <100> . The orientational

relation between the _- and _ phases which was established

differs from the familia_ common orientation of these phases

during martensite conversion [8].!

Preservation of the _,phase portions up to temperatures

(

7

|

I

significantly exceeding the temperature established previous-

ly for the conclusion _of martensite cenversion_is in agreement

with the results ef dilatometric and m_crostructural studies

in which it was shown that a certain part of the d--phase is

converted into the ?_,phase by the diffusion mehhanism [5,9,10].

Three portions can be isolated on the temperature differential

curve, obtained 9or slowly heated samples of a Fe-32.5% Ni

alloy [9]. Two of them are related to absorption, and one in-

volves heat loss• Th_ authors explain the course of the curve. .

as follows The:.,abso_ption found on the first, portion of the

curve (UP to a temperature of approximately 440°C) is connectedi

with the conversion of approximately 60% of the martensite into

IL.................austenlte. Then the reverse conversion does not occur during/ll9

heating to a temperature of 475°C. In this temperature range

440 - 475%) the authors assume that diffusion of iron and _ickel

occurs across the inte_phase boundary between martensite and

austenite. Here the martensite sheet combines with the nickel,

the d-phase is stabilized, and no structural changes are ob-

served. Above 475eC (the third portion of the curve) a reverse

conversion begins again which has, however, the characteristic{

feature_ of a diffusion conversion:

l) a surface ghost is not formed;

2) the conversio_ continues for 30 hr at 500eC;

3) the st_uct_reiwhich appeared when the austenite regions _.

were heated does net differs?from the structure of the portions

due to the diffuslom growth mechanism.

On the dilatomet_ic curves obtained for _ Fe_34% Ni alloy

[10] a g_adual decrease in the length of the sample was observed

between 200 and 280°C_ which is connected, the authors believe,

with the controlled diffusion process. Above _280°C the conver-

slon occurs in accordance with the martensite mechanism.

In the allo_ stud_ed by us, the basic part of the 4._phasei

8

I

is convertedinto the Y_phase by the martenslte mechanlsm. /119

However, the concluslan of the reverse martenslte conversion

is in a temperature i_terval where the mobility of the atoms

becomes somewhat grea_er so that conversion can proceed by

the martensite mechanism. Subsequent conversion takes place

by the slow-actlng diffusion mechanism. The time during

which the samples were kept in the oven in the experiments

performed proved to be insufficient for completing the _

conversion by this mechanism even at 700 and 800°C tempera-

tures. Moreover, in Connection with the nickel d_ficiency

the _esldues of the m_rtensite phase are converted into the .....

austenlte phase at higher temperatures.

The presence of two reverse _VJ_ , conversion mechanisms

iniron-nlckel alloys'is apparently the reason why the ghost

for reverse conversion does not always agree with the ghost

for a direct martensi_e conversion Eg, ll]. This is probably

involved in the observation that the directions I1_Jw and

[I_I_, are sometimes not parallel (cf. Figure 7).

conversion due to theThe possible occurence of a_?_

two mechanisms must bg taken into account when subjecting iron-

nickel alloys to heat !treatment. The role of the diffusion

mechanism is strengthened when the nickel content in alloys _

is lowered and the rate of heating is reduced. In this re- _

gard a substantial difference between the structure of the

internal and surface _ayers may arise during the hardening of

large test specimens.

9

f

Referenees

1. Grldnev, V° N.: Z_._rnal tekhntoheskoy fiziki, 1941, ll,1226. _ . /120

2. GolovchinerYa. M.: In the book: Probiemymetallovedenlyai flziklmetallova (Problemsin Metallurgy and in thePhysics of Metalls),Metallurguzdat,Moscow, 1951, 119.

3. KurdyumovG. V.: In the book: ibid.,Moscow, 1949, 133;Y_v_nlyazakalkl_iotpucka stall (The Phenomenonof Har-dening and the _emperlngof Steel), Metallurglzdat,Moscow, 1960.

4. Krauss, G. Acta Met. 1963, 11, 6.

5. Allen N. P., Earley C. C. JISI, 1950, 166, 4.

6. Baln E.C. Trans.AIME, 1924, 70, 25.

7. Bogers A. J. Acta Met., 1962, i0, 3.

..... 8. Nishiyama Z."Sci. Report Toh_ku Imp. Unlv. 1934-35,I, 23, 637.

9. Kessler H. Pitsch W. W. - Acta Met. 1967, 15, 2, 401.

I0. Jana S., Wayman C._M. -....Trans. AIME, 1967, 239, 8.

ii. Gorbach V. G., Butakova, E.D.: Fizika metallov i metallov-16_edeniye, 1963, ____,2.

10