Sunteți pe pagina 1din 6

Physica B 299 (2001) 236–241

Phase transformations in Fe–Ni system at mechanical alloying


and consequent annealing of elemental powder mixtures
S.D. Kaloshkina,*, V.V. Tcherdyntseva, I.A. Tomilina,
Yu.V. Baldokhinb, E.V. Shelekhova
a
Department of Physical Chemistry, Moscow State Steel and Alloys Institute, Leninsky prosp. 4 Moscow 117936, Russia
b
Institute of Chemical Physics Academy of Science, Kosygina str. 4, Moscow 117334, Russia

Abstract

Fe100xNix alloys (105x590 at%) were prepared by mechanical alloying (MA) of elemental powders in a high-
.
energy planetary ball mill and studied by X-ray diffractometry and Mossbauer spectroscopy. It is shown that the
concentration ranges of single-phase solid solutions of MA samples extend significantly as compared with those
obtained by conventional techniques. In our case, the BCC phase exists in the range from 0 to 20 at% Ni and FCC
phase from 30 to 100 at% Ni. Block size was 10–15 nm. Consequent annealing of MA samples resulted in further
extension of FCC single-phase concentration range to the relatively low Ni content (20 at%). This was caused by
considerable retardation of austenite–martensite transformation in MA alloys. The FCC alloys with 20–28 at% Ni were
found to be non-ferromagnetic at room temperature; only the paramagnetic component was observed in the
.
corresponding Mossbauer spectra. However, the treatments of low-nickel austenite alloys like cooling in liquid nitrogen
or mechanical deformation provoked austenite–martensite transformation and led to the rise of ferromagnetic
properties. # 2001 Elsevier Science B.V. All rights reserved.

.
Keywords: Fe–Ni alloys; Mechanical alloying; Martensite transformation; Mossbauer spectroscopy

1. Introduction structure and properties of mechanically alloyed


(MA) Fe–Ni alloys were investigated [1–7]. Recent
Fe–Ni system has attracted considerable atten- studies of MA binary alloys showed that the
tion as a basis for a number of alloys with special single-phase concentration ranges extend signifi-
magnetic and mechanical properties. Severe plastic cantly as compared with those for alloys produced
deformation of these alloys in various kinds of by conventional techniques [3,7–12]. For the
high-energy mechanoactivators enables one to Fe–Ni system also this tendency takes place, but
obtain alloys with new structures and properties. the results obtained by different researchers
Therefore, several studies were performed, where although similar in general do not coincide
completely [1,3,6]. In particular, it was observed
*Corresponding author. Tel.: +7-95-230-4667; fax: +7-95-
that the concentration ranges of the BCC
247-6001. and FCC single-phase solid solutions depend on
E-mail address: ksd@phch.misa.ac.ru (S.D. Kaloshkin). the milling intensity and shift to low nickel

0921-4526/01/$ - see front matter # 2001 Elsevier Science B.V. All rights reserved.
PII: S 0 9 2 1 - 4 5 2 6 ( 0 1 ) 0 0 4 7 3 - 2
S.D. Kaloshkin et al. / Physica B 299 (2001) 236–241 237

concentration at an increase in the milling intensity 3. Results and discussion


[1,3]. Here, some peculiarities of phase transfor-
mations in Fe–Ni alloys obtained in an AGO-2U MA Fe100xNix (105x590 at%) alloys were
high-energy mechanoactivator will be considered. studied both by X-ray diffractometry and Moss- .
One specific feature of the Fe–Ni system is that bauer spectrometry. As BCC and FCC phases can
the annealing of MA alloys may lead to a further be easily distinguished by these methods, they gave
extension of FCC single-phase concentration well-correlated results. As-milled samples with up
range even compared with that for the as-milled to 20 at% Ni contained single BCC phase, all the
alloys. The stabilization of FCC phase is caused by samples with Ni content higher than 28 at% Ni
the dependence of the austenite–martensite trans- contained only FCC phase, and the samples with
formation temperature on the structure para- 22, 24, 26, and 28 at% Ni contained both BCC and
meters. It is known that the refinement of grain FCC phases.
structure by rapid quenching of Fe–Ni alloys from Our results agree with the published data, espe-
melt [13] or by the MA processing [1] suppresses cially with the data reported in Ref. [3] (Fig. 1).
this non-diffusion transformation and low-nickel Actually, the milling intensity has a complex effect
single FCC phase may be obtained after annealing on the final structure of alloy. Two main para-
at temperature region of the austenite phase. Such meters change with increase of the intensity: on the
alloys are metastable at room temperature and the one hand, concentration of different types of
martensitic transformation in them can be induced structure defects increases and on the other hand,
by different treatments: the cooling [14,15], the the average temperature of the milling process
deformation [14], application of high magnetic increases. According to Ref. [3], the structure
field [1,16] or radiation [5]. Phase composition and defects, which are caused by the MA, lead to a
magnetic state of MA Fe–Ni alloys in the range of lesser destabilization of the FCC phase than of the
20–28 at% Ni are the subjects of our special BCC phase. Moreover, as it was mentioned in
interest. Ref. [1], according to the Fe–Ni phase diagram, an
increase of the milling temperature will stabilize
FCC and destabilise the BCC structure. That is
2. Procedure why, in general, an increase of the milling intensity

Pure carbonyl Fe (powder, 99.95%) and elec-


trolytic Ni (powder, 99.5%) were used as the
starting materials. Mechanical alloying of Fe–Ni
powder mixtures was carried out in an AGO-2U
planetary ball mill with water-cooled vials at a
milling intensity of about 20 W/g. The powder
mixture (15 g) and steel balls 8 mm in diameter
(150 g) were loaded into steel hermetic vials with a
volume of 160 cm3. All of the samples were milled
for 1 h; note that the alloy formation in Fe–Ni
system is completed in 30 min. The chemical
composition of MA alloys was controlled by local
X-ray microanalysis method, deviations of the
chemical compositions from the starting ones did
not exceed 1 wt%. Average oxygen contamination
determined by vacuum-melting method was about
1 wt%. The magnetic structure and phase compo-
.
sition were studied by Mossbauer spectroscopy Fig. 1. Phase boundaries for stable and obtained by MA
and X-ray diffractometry. technique Fe–Ni alloys.
238 S.D. Kaloshkin et al. / Physica B 299 (2001) 236–241

will extend the FCC concentration range and the side of low nickel content (Fig. 2). One can see
narrow the BCC concentration range (Fig. 1). further stabilization of FCC phase in the annealed
As it was shown earlier [12,17,18] the MA MA alloys. The transformation of MA Fe100xNix
treatment will always cause the narrowing of two- alloys (x>10) into FCC state during the annealing
phase concentration range in the case of binary then should be followed by the FCC ! BCC phase
systems, because of the constraint of the simulta- transformation in the cooling down to the room
neous existence of several phases with different temperature. However, this non-diffusion auste-
compositions. The only one phase with the lowest nite–martensite transformation appeared to be
Gibbs energy commonly forms at low temperature significantly suppressed for the Fe–Ni alloys
MA process. The AGO-2U mill, which is used prepared by the MA technique. It is interesting
here, combines very high intensity of milling (more that the alloys with 22, 24, 26 and 28 at% of Ni
than 10 times higher than that used in Refs. [1,3]) being single-phase FCC structured after such
with relatively low temperature inside the reactive treatment were paramagnetic, and the correspond-
space due to water cooling of the vials. This creates .
ing Mossbauer spectra of these alloys include only
such milling conditions, which allowed one to ob- one paramagnetic component (single line). The
tain rather narrow two-phase concentration range formation of the non-ferromagnetic alloys seems
(8–10%) in MA Fe–Ni system, as compared with very unusual for the Fe–Ni system. Fig. 2 illus-
not only equilibrium alloys but also with that trates the reasons for such behaviour. The MA
obtained earlier by the MA technique (Fig. 1). process with consequent annealing allowed one to
Thus, the tendency to form narrow two-phase decrease the maximum concentration of Ni at
concentration ranges becomes stronger at a de- which the BCC phase still exists, down to 20 at%
crease of the temperature and an increase of the and all the alloys with a higher Ni content had
intensity of MA process. only FCC phase. However, the dependence of the
After annealing of the samples at 6508C, the Curie temperature on the concentration for the
concentration ranges of phases existence shifted to FCC phase intersects the room temperature at
about 28 at% Ni. As a result, the alloys with
22–28 at% of Ni were paramagnetic.
It is known that the temperature of austenite–
martensite transition is rather sensitive to the
structure transformations of alloy [13,14]. This
transition point should depend on the grain size,
the presence of various types of defects, etc. be-
cause of the closeness of Gibbs energies for BCC
and FCC phases. As it was mentioned before, the
grain refinement in Fe–Ni alloys at rapid quench-
ing (RQ) from melt [13] or MA treatment [1] leads
to a decrease in the temperature of martensite
formation. Therefore, a strong effect of the MA
method of samples preparation, which is used
here, on martensitic point is also expected.
The dependences of austenite-martensite trans-
formation temperature on the composition are
shown in Fig. 3 for the Fe–Ni alloys produced by
different ways: (1) as-cast and quenched from solid
state [15], (2) rapidly quenched (RQ) from melt
Fig. 2. Concentration ranges and the corresponding depen- [13], and (3) obtained by the MA technique [1].
dences of Curie temperatures of BCC and FCC phases in Fe–Ni The curve corresponding to the data of this work
alloys: as-milled and after annealing at 6508C. (4) also is plotted in Fig. 3. As one can see, the
S.D. Kaloshkin et al. / Physica B 299 (2001) 236–241 239

Such significant difference of microstructure


data obtained in Ref. [1] and in this work, most
probably, is caused by the difference in the gas
(primarily oxygen) contamination of milling atmo-
sphere. In Ref. [1] special care has been taken for
protection of Fe–Ni powders at MA processing
against oxidation: it was reported that as-milled
samples are the coarse particles; this confirms a
low oxygen content in them. On the contrary, in
the present work, the content of oxygen in the
samples was rather high (about 1 wt%), though
no oxides were observed. Correspondingly, the
produced MA samples looked like rather fine
powders. We suppose that the significant oxygen
impurity may also cause much stronger refinement
of alloy microstructure as compared with oxygen-
free samples.
Preparation of Fe–Ni alloys in the range of
20–30 at% Ni with a step of 2% Ni allowed us to
clarify the range of single FCC phase existence at
room temperature. Fig. 4 gives the X-ray patterns
Fig. 3. The dependences of austenite–martensite transforma- .
and Mossbauer spectra for the most interesting
tion temperature on the composition of the Fe–Ni alloys: (1) as- composition with 22 at% Ni after different treat-
cast and quenched from solid state [15], (2) rapidly quenched
ments. Annealed alloy was completely paramag-
(RQ) from melt [13], (3) obtained by MA technique in Ref. [1],
and (4) the data of this work. netic, only single line corresponding to g-phase was
observed (Fig. 4a). Alloys with 24 and 26 at% Ni
had similar spectra. For alloys with 28 at% Ni the
single line was considerably wider than for other
strongest depression of the martensite transforma- alloys. This indicates that the Curie point for this
tion was achieved when the RQ technique was alloy is close to the room temperature.
used; the MA technique also allowed one to A subsequent cooling and mechanical treat-
decrease the transition temperature, but to a lesser ments brings up further evolution of the phase
extent. It is interesting that the depression of composition and magnetic structure of the alloys.
martensite transformation, which is revealed in Cooling of the samples in liquid nitrogen (Fig. 4c)
this work, is stronger than that observed earlier [1], stimulated austenite–martensite transformation.
when a similar MA technique was used. We may Partial austenite–martensite transformation in
compare the values of average block size of FCC these alloys can be evoked even by low-intensive
phase calculated from X-ray patterns for as-milled deformation of the powder alloy in a mortar
Fe60Ni40 composition: about 40 nm in Ref. [1] and (Fig. 4d). Since such transformation is followed by
10 nm in this work. Of course, after annealing of an appearance of ferromagnetism of the samples,
samples at 6508C we found an increase of the this effect seems to be interesting for practical use,
block size to 48 nm for the same alloy, but it still for example, in deformation sensors.
remained in the nanometric range. For the most After long-term annealing of the MA samples,
interesting concentration range of 20–30 at% of Ni their austenite–martensite transition temperature
the block size varied from 35 to 50 nm. This means approaches that for conventionally prepared
that in our case we obtained the block sizes after Fe–Ni alloys. It was shown that Fe78Ni22 MA
annealing comparable to those for the as-milled alloy after annealing at 7008C for 12 h become
samples in Ref. [1]. martensite at room temperature (Fig. 5). It means
240 S.D. Kaloshkin et al. / Physica B 299 (2001) 236–241

Fig. 5. X-ray patterns of Fe80Ni20 alloy annealed for (a) 30 min,


(b) 3 h, and (c) 12 h at 7008C.

4. Conclusions

.
Fig. 4. X-ray patterns and Mossbauer spectra of Fe78Ni22 Fe–Ni alloys were prepared by MA technique in
alloy: (a) as-milled; (b) annealed at 7008C for 1 h; (c) cooled a high-energy planetary ball mill. MA samples
down to liquid nitrogen temperature; (d) pounded in mortar for showed a significant widening of concentration
5 min. (m) X-ray patterns of martensite, (au) X-ray patterns of
ranges of single-phase solid solutions compared
austenite.
with alloys obtained by conventional methods.
Subsequent annealing of MA samples showed
widening of FCC single-phase concentration
that the very mechanical treatment, but not range down to 22 at% Ni. It was found that the
chemical impurities, primarily affects the auste- austenite–martensite transition temperature de-
nite–martensite transformation. The mechanism of creases significantly for mechanically alloyed
this effect is not completely clear now. It seems Fe–Ni powders. MA alloys with 22–28 at% Ni
that the main cause of this transformation after annealing in austenite temperature region
suppression is the small block size in the MA have a single phase with FCC structure and are
FCC structure; however, the structure defects like not ferromagnetic at room temperature. Low-
dislocations, stacking faults and their interaction temperature treatment as well as deformation
also may play an important role. leads to partial transformation of FCC phase into
S.D. Kaloshkin et al. / Physica B 299 (2001) 236–241 241

BCC phase at 22 at% Ni. The refinement of [7] V.V. Tcherdyntsev, S.D. Kaloshkin, I.A. Tomilin,
microstructure of austenite phase is considered as E.V. Shelekhov, Yu.V. Baldokhin, Nanostruct. Mater. 12
(1999) 139.
one of the most important factors resulting in the
[8] N.S. Kohen, E. Ahlswede, J.D. Wicks, O.A. Pankhurst,
depression of martensite transformation. J. Phys: Condens. Matter 9 (1997) 3259.
[9] L. Schultz, J. Less-Common. Metals 145 (1988) 233.
[10] V.V. Tcherdyntsev, S.D. Kaloshkin, I.A. Tomilin, E.V.
Acknowledgements Shelekhov, Yu.V. Baldokhin, Z. Metallk. 90 (9) (1999)
747.
[11] S.D. Kaloshkin, I.A. Tomilin, G.A. Andrianov, U.V.
This work was supported by INTAS Project No. Baldokhin, E.V. Shelekhov, Mater. Sci. Forum 235–238
99-01741. (1997) 565.
[12] S.D. Kaloshkin, J. Metastable. Nanostuct. Mater. 8 (2000)
591.
References [13] Y. Inokite, B. Cantor, J. Mater. Sci. 12 (1977) 946.
[14] G.V. Kurdjumov, L.M. Utevskii, R.I. Entin, Transforma-
tions in Iron and Steel, Nauka, Moscow, 1977.
[1] C. Kuhrt, L. Schultz, J. Appl. Phys. 73 (1993) 1975. [15] M. Hansen, K. Anderko, Constitution of Binary Alloys,
[2] C. Kuhrt, L. Schultz, J. Appl. Phys. 73 (1993) 6588. McGraw-Hill, New York, 1985.
[3] L. Hong, B. Fultz, J. Appl. Phys. 79 (1993) 3946.
[16] K. Shimizu, T. Kakeshita, ISIJ Int. 29 (1989) 97.
[4] R.B. Scorzelli, Hyperfine Interactions 110 (1997) 143. [17] E. Ma, H.W. Sheng, J.H. He, P.H. Schilling, Mater. Sci.
[5] T.M. Lapina, V.A. Shabashov, V.V. Sagaradze, Eng. A 286 (2000) 48.
V.L. Arbuzov, Mater. Sci. Forum 294–296 (1999) 767. [18] P.J. Schiling, J.H. He, R.C. Tittsworth, E. Ma, Acta
[6] Yu.V. Baldokhin, V.V. Tcherdyntsev, S.D. Kaloshkin,
Mater. 47 (1999) 2525.
G.A. Kochetov, Yu.A. Pustov, J. Magn. Magn. Mater.
203 (1999) 313.

S-ar putea să vă placă și