alfe2

7/28/2019 alfe2

1/13

7/28/2019 alfe2

2/13

diagram. For the isothermal sections shown in the present pa-per the data from the assessment of Murray [28] have beentaken.

2.2. Al e Ti system

Following the elaborate assessment by Murray [29] theAle Ti system has been updated regularly [30e 32]. Only re-cently a new comprehensive assessment of the Al e Ti phasediagram has been performed [33] and the reader is referredto that publication for any details. The phase diagram derivedfrom this assessment is shown in Fig. 1. Compared to the latestupdate of the Al e Ti phase diagram [32] it differs especially inthe Al-rich part, as the assessment takes into account resultsfrom a number of recent papers, which have not yet been cov-ered in the update. It is noted that none of the CALPHAD-typedescriptions based on the available thermodynamic data [34e47] matches the assessed phase diagram as a whole, though

remarkable agreements between individual equilibria of theassessed phase diagram and the calculated ones exist.

2.3. Al e Fe system

Since the assessment of the Fe e Al system by Kattner [48](Fig. 2) numerous new results have been obtained within thissystem. Many of them concern details of the B2 and D0 3orderings and the variation of the vacancy concentration asfunction of temperature and composition in the Fe-rich partof the system. As they are too numerous to be dealt with in de-tail within the scope of the present assessment, discussion willfocus on selected works only.

The phase equilibria between a Fe (A2), FeAl (B2), andFe3Al (D03) have been re-investigated by in situ high-temper-ature neutron diffraction [49] and by investigations of diffu-sion couples [50]. In the latter study it was found that thea Fe/ a Fe Fe3Al phase boundary is actually at higher Fecontents below 450 C than suggested in the assessment byKattner [48]. The phase equilibria between the bcc phases

have also been calculated by the Monte Carlo (MC) method[51], MC and cluster variation (CVM) methods [52], and theKirkwood method [53]. Compositions of the Al-rich interme-tallic phases at 1000 C have been experimentally determinedin Refs. [54,55] .

The liquidus in the Al-rich part has been investigated by

differential thermal analysis (DTA) [56]. It has also been in-vestigated by differential scanning calorimetry (DSC) and sub-sequently the stable and metastable Al-rich parts have beencalculated [57]. The solubility of Fe in liquid Al has been cal-culated by Pashechko and Vasyliv [58].

Data for the crystallographic structures of Fe 2Al5 [59,60]and Fe4Al13 (FeAl 3) [61,62] have been published. The crys-tallographic structure of the high-temperature phase 3 is stillnot settled [63] as this phase decomposes spontaneouslyduring quenching. Results from ternary Fe e Ale X alloys in-dicate that it may have the hexagonal Al 8Cr5-type structure[64e 66].

3. Effect of alloying the binary phaseswith the third component

A prominent feature of the Al e Fee Ti system is the largesolid solubility of Al in the binary Fe e Ti phases. In theC14-type Laves phase Fe 2Ti (crystallographic data of allphases are given in Table 1 ) more than two-thirds of Fe canbe substituted by Al [67]. Al substitutes for Fe on both of the two different crystallographic Fe positions alike and notransformation to another Laves phase polytype has been re-ported. The substitution of Fe by Al leads to an increase inthe lattice parameters and their dependence on compositionhas been determined [67,68] .

In B2-ordered FeTi Al can also substitute for Fe to a largeextent. At 1300 C a composition of Fe 5Al42 Ti53 has been re-ported [69] (all compositions are given in at.% throughout thispaper). Apparently the B2 order is not affected by this substi-tution. The dependence of the lattice constant on compositionhas been determined in Refs. [67,70] . In both phases, Fe 2Ti

0 10 20 30 40 50 60 70 80 90 100

T e m p e r a

t u r e

( C )

600

700

800

900

1000

1100

1200

1300

1400

1500

1600

1700

1200 C

1120 C

Ti

Ti Ti 3Al

1170 C TiAl

Ti

1690 C

1491 C 1456 C

L

1412 C

1215 C

1d-APS

TiAl 3(l)

TiAl 3(h)

TiAl 2

977 C

665 CAl

at. Al

Fig. 1. The Ale Ti system according to the most recent assessment [33].

ARTICLE IN PRESS

2 M. Palm, J. Lacaze / Intermetallics xx (2006) 1 e 13

+ MODEL

7/28/2019 alfe2

3/13

and FeTi, the solid solubility for Al increases markedly withincreasing temperature [67].

Disordered bTi (A2) as well has a marked solid solubilityfor Al of about 25 at.% [71]. Kainuma et al. [71] found thatbTi becomes ordered B2 by progressive substitution of Ti byFe Al. At 1000 C they found a continuous range of solidsolutions between bTi (A2) and FeTi (B2) with a second orderdisorder e order transition from A2 to B2 at about 70 at.% Ti.Though, the course of the tie-lines on the Ti-side in Ref. [71]suggests the existence of a three-phase eld A2 B2 Ti3Al

at this temperature, i.e. the existence of a rst-order transitionwith a narrow A2 B2 two-phase eld. Nevertheless the A2/ B2 transition may change from rst to second order at some-what higher temperatures.

The solid solubility for Fe in all Al e Ti phases is very lim-ited. The maximum content of Fe in a Ti is about 1 at.% at anAl content of 44 at.% [72]. For more Ti-rich compositions thesolid solubility for Fe in a Ti is even more restricted. ForTi3Al, TiAl, TiAl 2 and TiAl 3 the solid solubility for Fe rangesbetween 1.2 and 2.5 at.% [67,73] and varies little with tem-perature [67] (cf. Table 1). The site occupancies in TiAland Ti 3Al alloyed with Fe have been experimentally deter-mined by atom location channelling enhanced microanalysis(ALCHEMI) [74].

g Fe has only a very limited solid solubility for both Al andTi. The solid solubility for Ti in a Fe (A2) and FeAl (B2) in-creases with increasing temperature and both phases containup to about 10 at.% Ti at 1000 C [67]. The site occupancyof Ti in B2-ordered FeAl has been determined by ALCHEMI[75] and has been modeled [76,77] . D03-ordered Fe 3Al caneven contain up to 25 at.% Ti in solid solution [67,78] .

By alloying Fe-rich Fe e Al alloys with Ti the transitiontemperatures for the D0 3 /B2 and B2/A2 transitions increasemarkedly [8,13,78 e 84]. The D03-type ordering is maintainedbecause Ti substitutes in Fe 3Al for Fe on , the 4(b)

Wyckoff site of the lattice [85,86] . Ohnuma et al. [78] studied

in detail the dependence of the transition temperatures oncomposition. They found that the D0 3-structure is stable be-tween binary Fe e 25 at.%Al and Fe e 25 at.%Al e 25 at.%Tiand that the D0 3 /B2 transition temperature simultaneously in-creases from 547 C in the binary to 1212 C in the ternarysystem. No clear distinction has been made up to now whenthis phase is denoted as D0 3 (binary) or L2 1 (ternary, Heus-ler-type). The lattice parameter of D0 3 increases by the substi-tution of Fe by Ti and its dependence on composition has beendetermined in Refs. [84,87] . An increase of the lattice param-

eter was also observed with increasing Ti content alongFeAl (1 x )Ti x [88].

Because the binary compound 3 decomposes by a eutectoidreaction during quenching to room temperature its crystallo-graphic structure has not been determined yet. Also as-cast ter-nary Ale Fee Ti alloys with compositions near this binaryphase show similar ne-scaled microstructures, which presum-ably were generated by eutectoid decomposition of 3 duringcooling [66]. EDS analyses of some unaltered grains revealedthat they contained at least 6.5 at.% Ti. XRD indicates thatthese grains have the hexagonal Al 8Cr5-type structure and pro-visional lattice constants have been established asa 0 1.268(4) nm and c0 0.790(3) nm [66]. It is assumedthat this is not a ternary phase but that this composition lieswithin the ternary solid solubility range of the binary phase 3.

FeAl 2 and Fe2Al5 both have limited solid solubilities for Tiof 1.8 and 2.5 at.%, respectively [67]. Fe4Al13 can dissolveabout 6.5 at.% Ti [67]. A structure renement of Fe 4Al13 con-taining about 5.5 at.% Ti revealed that Ti replaces Al only oncertain crystallographic sites [89].

4. Ternary phases

The existence of two ternary intermetallic phases has beenconrmed. The phase t 2 is stable over a wide range of compo-

sitions and, depending on composition, exists in two polytypes

Fig. 2. The Fee

Al system according to the assessment by Kattner [48] .

ARTICLE IN PRESS


+ MODEL

7/28/2019 alfe2

4/13

7/28/2019 alfe2

5/13

the presence of t 2* in quenched samples, but rst results of insitu neutron diffraction experiments failed to prove the exis-tence of the tetragonally distorted t 2* polytype at 800 C,where instead the cubic t 2 polytype was found [91].

The cubic phase t 3 with L1 2 structure is stable at composi-tions of 7.5 at.% Fe, 63.9 e 66.6 at.% Al and 28.6 e 25.8 at.%

Ti, and no marked broadening of the single-phase eld isobserved with temperature between 800 and 1000 C [67].Data available for 1200 C indicate that at this temperaturet 3 is stable within a wider range of compositions [92]. Sitepreferences in t 3 and the long-range order parameter havebeen determined in Refs. [16,93,94] .

The existence of the phases Fe 2AlTi ( t 1 , Heusler-typephase with L2 1 structure [95]) and Fe25 Al69 Ti6 (X, "low sym-metry" [68]) has to be ruled out [22]. Both phases are appar-ently no phases on their own. While t 1 corresponds to themaximum solid solubility of Ti in Fe 3Al, Fe25 Al69 Ti6 mostlikely has been confused with the extended homogeneity rangeof Fe4Al13 .

Three more phases have been described recently, thoughnot enough details are known right now to decide whetherthese are stable ternary phases. Two of these phases actuallymay not be true ternary compounds, but may form continuoussolid solutions with binary compounds. This point has notbeen settled because the individual homogeneity ranges of the phases in question have not been established yet as func-tion of temperature and composition.

At a composition of 60.5 Al, 33.0 Fe, 6.5 Ti Ducher et al.[66] detected a hexagonal phase of Al 8C5 type. It is assumedthat this is actually not a ternary phase, but represents the ter-nary solid solubility range of 3 (see also above).

Ducher et al. [96] found extra diffraction spots in TiAl thathas been alloyed with Fe. The diffraction pattern could be ex-plained in that by adding Fe the stacking sequence in {111} L10is modied. Though the original description of this phase ac-tually came from a quinary alloy, the same type of orderingwas later on found in an as-cast ternary alloy at a compositionof 46.5 at.% Al, 2.5 at.% Fe, 51.0 at.% Ti [20]. Whether this isa stable phase and if it forms continuously from TiAl or if it isseparated from TiAl by a two-phase eld has not been estab-lished yet.

Levin et al. [69,97 e 99] studied the evolution of the micro-structure in an alloy containing 48.5 at.% Al, 1.9 at.% Fe,49.6 at.% Ti after various heat treatments. At a compositionof 38 at.% Al, 10 at.% Fe and 52 at.% Ti they observed a phasewith a tetragonal unit cell of a 0 1.15 nm, c0 1.38 nm. Thephase was found in samples which were water-quenched afterannealing at 1400 and 1200 C [69,98] , but not in sampleswhich were slowly cooled from this temperatures or water-quenched from 700 to 1070 C [99]. The phase has beendenoted as t 02 , as it had been originally considered to be thetetragonal polytype of the t 2 phase [69]. Subsequent investiga-tions revealed that t 02 forms during cooling from Al-richB2-ordered FeTi [98]. That t 02 may be a metastable phase isindicated by the fact that the compositions of the other phasesmeasured in a t 02-containing sample that has been water-

quenched from 1200 C [98] differ from those reported by

Kainuma et al. [72] for the same composition range. Still de-tails on the stability of the three phases discussed above needfurther clarication.

5. Liquidus surface

The basic layout of the liquidus surface has been estab-lished by Seibold [95] and though further amendments havebeen made [22,24] even the latest of these versions had beenconsidered as tentative. Therefore, Ducher et al. re-investi-gated the liquidus surface by DTA and by performing opticalmicroscopy, X-ray diffraction (XRD), and energy dispersiveanalyses on a scanning electron microscope (SEM e EDS) onas-cast samples [66]. The liquidus projection shown inFig. 3 is based on that study. Compared to the last e tentativee version by Raghavan [24] it differs mainly in that there is noprimary eld of solidication of the Ti-rich polytype of t 2 andno continuous solid solubility between bTi and FeTi exists upto the liquidus. Actually, t 2 was found to decompose at1075 C by the peritectoid reaction: TiAl FeTi Fe2Ti 4t 2 (Ti-rich) (Pd 1) as mentioned above. Otherwise there aresome changes regarding the positions of invariant lines.Most prominent is the nding that the area of primary solidi-cation of 3 extends up to the eutectic trough.

6. Invariant reactions and reaction scheme

Reaction schemes of the Al e Fee Ti system have been pre-viously presented [22,24,26,66,67,95] and a discussion on var-ious sections of the reaction scheme has been provided byDucher [20]. The reaction scheme shown in Fig. 4 is based

on the recent one by Ducher et al. [66]. Amendments have

80

Ti

AlFe

Fe

Fe 2Ti

2*3

TiA l3

Fe 4Al 13Fe 2Al 5

TiAl

TiFeTi

80

60

40

2080

60

40

20

Al

*Ti

1397

1338 1303

14081356

1248

1116

1101

13031255

13011331

1326

13011322

1395

1300

1239

1286

1325

1287

1 2 0 0

1 3 0 0

1 4 0 0

1 2 0 0 1

3 0

0

1 4 0 0

1 3 0 0

1 4 0 0

1 2 0 0

1447

1222

1360

1296

1235

1371

1323

1 6 0 0 1 5 0 0

1 5 0 0 1 1 0 0 1 0 0 0

e1 U6 U7 U9E1

P3 U10

e5U18

p6

P2 P1

U2M2

U3U4 U1

p4

e4

p5 e2 e3

p3

p1

p2

Al (at. % )

F e ( a t . %

) T i ( a t . %

)

M1

604020

Fig. 3. Liquidus projection of the Al e Fee Ti system based on Ref. [66] . Dotscorrespond to alloys for which the indicated liquidus temperatures (in C)have been determined by DTA. Single and double arrows designate peritecticand eutectic reactions, respectively. Invariant reactions are labelled accordingto the reaction scheme shown in Fig. 4. Isotherms are tentative.

ARTICLE IN PRESS


+ MODEL

7/28/2019 alfe2

6/13

Fig. 4 (continued ).

ARTICLE IN PRESS


+ MODEL

7/28/2019 alfe2

7/13

been made in that data for invariant reactions in the Al e Ti sys-tem have been updated from Ref. [33]. From this update it isclear that the phase Al 5Ti2 does not exist as such, but is a partof a series of one-dimensional antiphase domain structures(1d-APS). It is now conrmed that these structures are not sep-arated from TiAl by a two-phase eld above 1215 C [100e102], which implicates that they form either stably at hightemperatures or metastably during cooling from Al-rich TiAlby ordering of excess aluminium. To take these new ndingsinto account, the reactions TiAl 4 TiAl Ti2Al5 (c1) andTiAl Ti2Al5 4 TiAl 2 t 3 (U5) from Ref. [66] have been

replaced by TiAl/1d-APS 4 TiAl 2 (c1) and TiAl/1d-

APS 4 TiAl 2 t 3 (C1), respectively, and the reactionTiAl Ti2Al5 4 TiAl 2 (pd1) at 1215 C has been removed.Also some errors have been corrected. The peritectic reactionL t 2* t 3 4 Fe4Al13 at z 1100 C is now correctly de-noted as P3 (instead of E 3), the non-existing reactionL t 3 4 Fe2Al5 t 2* (U10 ) has been removed and the tem-perature for L t 3 4 TiAl 3 Fe4Al13 has been changed toz 1070 C [20].

It is noted that temperatures given for most solid-state reac-tions above 1000 C are estimates, which need further study tobe accurately established [66]. Ternary invariant reactions, for

which temperatures have been established, e.g. by DTA, are

Fig. 4. Reaction scheme of the Al e Fee Ti system. The scheme is based on the one by Ducher et al. [66]. Data for invariant reactions in the Al e Ti system have beenupdated from Schuster and Palm [33]. The following designations for the phases are used: t 2 , Ti-rich variant; t 2* Al-rich variant; a Ti, low temperature(< 1170 C); a *Ti, high temperature ( > 1120 C); LAl/Fe : two separate melts which exist below M 1 . The different types of ordering in a Fe (A2, B2, D0 3 /L21),distinction between TiAl 3 (h) and TiAl 3 (l), and the possible merging of the phases FeTi and bTi at high temperatures are not taken into account (see text). Re-actions are labelled as e/E: eutectic, p/P: peritectic, ed/Ed: eutectoid, pd/Pd: peritectoid, c/C: congruent, U: transition, M: maximum with small and capital lettersdenoting binary and ternary reactions, respectively. Bold boxes denote terminal reactions. Bold and underlined three-phase equilibria are present in the isothermalsection at 800 C, the lowest temperature for which a complete isotherm exists.

ARTICLE IN PRESS


+ MODEL

7/28/2019 alfe2

8/13

TiAl FeTi Fe2Ti 4 t 2 (Pd1) at 1075 C [66], 3 4 a Fe FeAl 2 t 2* (Ed1) at 1041 C [66], and L TiAl 3 4 FeAl 3 Al at 658 C (U18 ) [103] .

In the reaction scheme in Fig. 4 the different types of order-ing in a Fe (A2, B2, D0 3 /L21) have not been taken into ac-count. Also no distinction between TiAl 3 (h) and TiAl 3 (l) is

made, as the details of the stability ranges of the two poly-morphs have not been yet sufciently established [33]. FeTi(B2) and bTi (A2) are treated as separate phases, though thetwo single-phase elds may merge at high temperatures [71].

The reaction scheme presented in Fig. 4 is in accordancewith the recent versions of the binary sub-systems, with theliquidus projection shown in Fig. 3 and the isothermal sec-tions, which are discussed in the following section.

7. Isothermal sections

Complete isothermal sections have been determined exper-imentally for 800 C [67,68,95] , 900 C [19] and 1000 C[4,67] and in addition a number of partial isotherms havebeen determined. Table 3 summarises the experimental studiesof phase equilibria in the Al e Fee Ti system.

As there are a considerable number of investigations onphase equilibria in the temperature range 800 e 1000 C,which are at least to most parts qualitatively in accordancewith each other, complete isotherms for 800 C (Fig. 5a),900 C (Fig. 6) and 1000 C (Fig. 7a) have been established.

They are consistent with each other and the reaction schemeshown in Fig. 4.

The isotherm for 800 C (Fig. 5a) is based on the one byPalm et al. [67]. This isotherm had been experimentally deter-mined by metallography, XRD, and EPMA on quenched sam-ples (cf. Table 3). The common features and differences

between this isotherm and previous investigations by Markivet al. [68] and Seibold [95] have already been discussed [67].Since then, few new data have become available. Nwobuet al. [104] determined the phase equilibria between a Ti andbTi at 800 C and their results are in good agreement withthe isotherm shown in Fig. 5a. In contrast most of the data de-termined for Al-rich Al e Ti alloys with Fe contents up to 8 at.%Fe [105] are in disagreement with Fig. 5a. As these data appar-ently lead to the improbable phase diagram proposed by Yangand Goo [105] they have not been considered in Fig. 5a.Ohnuma et al. [78] investigated the ordering and phase separa-tion in the bcc phase in the Fe corner. Their results togetherwith those of other studies [13,20,79,80,82,83] are shown inFig. 5b. These results clearly show the existence of a two-phaseeld between a Fe (A2) and L2 1 , and the respective changeshave been made in Fig. 5a. The isotherms shown in Figs. 5a,6 and 7a have also been updated to comply with the recentassessment of the Al e Ti system [33].

Fig. 6 shows the isothermal section at 900 C. It has beenbasically adopted from Ref. [106] with adjustments for the bi-nary Fee Ti and Al e Ti systems, the composition of the ternary

Table 3Experimental studies of phase equilibria in the Al e Fee Ti system

Temperature ( C) Composition range (at.%) Number of alloys(heat treatment(s))

Experimentaltechniques

Reference

550, 800, 1000, 1100 Ti corner > 60 at.% Ti 30 alloys (500 C/500 h;800 C/200 h; 1000 C/100 h;1100 C/100 h)

LOM, XRD [123]

550, 800, 1100 Ti corner > 60 at.% Ti About 80 alloys (550 C/1000 h;800 C/200e 400 h; 1100 C/6e 75 h)

LOM, XRD [124]

550 Ti corner > 58 at.% Ti About 76 alloys (550 C/1000 h) LOM, XRD [125]700, 800 Ti corner > 80 at.% Ti ( a Ti bTi) 7 alloys (700 C/720 h; 800 C/240 h) LOM, SEM, (TEM), EPMA [104]800 Full isotherm > 100 alloys (800 C/600e 900 h) LOM, XRD [68]23, 800 > 50 at.% Ti at 23 C.

Full isotherm at about 800 Cabout 64 alloys (800 C/192 h) LOM, XRD [95]

800 Al-rich Al e Ti alloys < 10 at.% Fe 12 alloys (800 C/ 240 h) SEM, HR-TEM, (EPMA) [105]800, 900 A2 L21 , B2 L21 5 alloys, 2 diffusion couples (900 C)

(800 C/672 h; 900 C/336 h)

SEM, TEM, EPMA, DSC [78]

800, 1000 Full isotherms 64 alloys (800 C/500 h; 1000 C/100 h;6 diffusion couples 1000 C/500 h)

LOM, SEM, XRD, EPMA [67]

900 Full isotherm 30 alloys (900 C/500 h) LOM, XRD, EPMA [19]900 Ti 3Al TiAl 1e 3 at.% Fe 13 alloys (900 C/240 h) XRD [73]1000 (Full isotherm) supplement to Ref. [67] 10 alloys (1000 C/96 h) LOM, XRD, EPMA [4]1000 Emphasis on phase equilibria with FeTi 25 alloys (1000 C/500 h) LOM, XRD, EPMA [70]1000 Ti-rich part > 50 at.% Ti 14 alloys (1000 C/168 h; 2 diffusion

couples 1000 C/3 h)LOM, TEM, EPMA [71]

1000, 1200, 1300 Phase equilibria among a Ti,bTi, TiAl, Ti 3Al

5 alloys (1000 C/168 e 504 h;1200 C/168 h; 1300 C/24 h)

LOM, EPMA [72]

1000, 1150 Emphasis on phase equilibriawith Al 8FeTi 3 (L12)

14 alloys (1000 C/144 h; 1150 C/48 h)35 alloys (1000 C/48 h; 1150 C/24 h)

LOM, SEM, (TEM), XRD, EPMA [107]

1200 Phase equilibria with Al 8FeTi 3 (L12) 9 alloys (1200 C/500 h) LOM, SEM, (TEM, XRD), EPMA [92]

LOM: Light optical microscopy; SEM: Scanning electron microscopy; TEM: Transmission electron microscopy; XRD: Powder X-ray diffraction; EPMA: Electron

probe microanalysis; DSC: Differential scanning calorimetry; when given in brackets the technique has been employed but no specic results are reported.

ARTICLE IN PRESS


+ MODEL

7/28/2019 alfe2

9/13

melt [66], and the composition of FeTi at the three-phase equi-librium FeTi Ti3Al t 2 . The only additional data whichhave become available at 900 C since the previous study[106] are from Hao et al. [73] for the solid solubility of Fein Ti3Al and TiAl. Their data indicate a slightly higher solidsolubility for Fe in these two phases than shown in Fig. 6but are otherwise in full agreement with the phase equilibriashown in Fig. 6.

The isothermal section at 1000 C (Fig. 7a) is based on thework by Gorzel et al. [4], which was an updated version of a previous study [67]. As for this temperature a considerableamount of additional data is available now, several amend-ments compared to the original isotherm have been made inFig. 7a. Besides adjustments for the Fe e Ti and Ale Ti binarysystems, the composition of the ternary melt has been cor-rected according to Ducher et al. [66]. Phase equilibria inthe Ti-rich corner of the Al e Fee Ti system at 1000 C havebeen determined by Kainuma et al. [71]. Their compositionsmeasured for two-phase FeTi bTi t perfectly with thosefrom Palm et al. [67], while their other data supplement themissing phase equilibria in the Ti corner ( Fig. 7a). ThoughKainuma et al. found that bTi forms a continuous range of solid solutions with FeTi at 1000 C this feature has notbeen adopted in Fig. 7a for reasons discussed in Section 3.Consequently the single-phase areas of bTi and FeTi as deter-mined by Kainuma et al. [71] are separated by a narrow two-phase eld in Fig. 7a and the adjacent three-phase equilibriaare indicated by dashed lines. Mabuchi et al. [107] determinedpartial isothermal sections at 1000 and 1150 C around thephase t 3 . Qualitatively most of their results for 1000 C arein agreement with the isotherm shown in Fig. 7a, except thatthey nd the two-phase equilibrium TiAl 2 t 2* instead of Fe2Ti t 3 at this temperature. As the respective invariant re-

action involving these four phases takes place slightly below

Al0 10 20 30 40 50 60 70 80 90 100

Ti

0

10

20

30

40

50

60

70

80

90

100

Fe

0

10

20

30

40

50

60

70

80

90

100 B2A2

L2 1

800 C

FeTi

Ti

Ti

Ti 3Al

TiAl

T iAl 2

TiAl 3

2

2*3

F e 4 A

l 1 3

L

F e A l 2

F e 2 A l 5

A

0 10 20 30 40 50

10

20

Fe

80

90

100 B2A2

L2 1

800 C

binaryphase equilibria Palm et al. [67]B2 Fortnum&Mikkola [80]2-phase Mendiratta et al. [79]L21 Mendiratta et al.[79]

EPMA Mendiratta et al. [79]B2 Anthony&Fultz [82]B2 Wright et al. [83]L21 Wright et al. [83]

2-phase Ohnuma et al. [78]EPMA Ohnuma et al. [78]A2/B2, B2/L2 1 Ohnuma et al. [78]B2 Stein et al. [13]A2 Ducher [20]B2 Ducher [20]L21 Ducher [20]

B

Fig. 5. (a) Isothermal section of the Al e Fee Ti system at 800 C. Theisotherm is based on the one given in Ref. [67] . Phase equilibria in the Fe cor-ner have been modied according to the results shown in Fig. 5b. Dotted linesdenote second-order transitions and dots on the axis mark compositions of binary phases. (b) Partial isothermal section of the Fe corner of the Al e FeeTi system at 800 C. Phase relations between a Fe (A2), FeAl (B2), andFe2(Fe,Ti) 1Al (D03 /L21) have been examined by high-temperature XRD[80] , TEM and EPMA [79], DTA [20,82] , measurement of the resistivity asfunction of temperature [83], DSC, TEM, and EPMA [78], and TEM andDTA [13]. Solid lines denote rst-order transitions while dotted lines denotesecond-order transitions. Dashed lines depict phase equilibria from Palmet al. [67] .

Al0 10 20 30 40 50 60 70 80 90 100

Ti

0

10

20

30

40

50

60

70

80

90

100

Fe

0

10

20

30

40

50

60

70

80

90

100

900 C

FeTi

Ti

TiAl

3

F e A l 2

2*

2

L

Ti 3Al

T iAl 2

TiAl 3

Ti

B2

L2 1

A2

F e 4 A l 1 3

F e 2 A l 5

Fig. 6. Isothermal section of the Al e Fee Ti system at 900 C based onRef. [106].

ARTICLE IN PRESS


+ MODEL
http://-/?-http://-/?-

7/28/2019 alfe2

10/13
http://-/?-

7/28/2019 alfe2

11/13

with the reaction scheme in Fig. 4 as are the other tie-trianglesthat are shown in Fig. 8.

It is noted that also for temperatures below 800 C some in-formation on phase equilibria exists (cf. Table 3 ).

8. Magnetic and electrical data

Some magnetic properties of Al e Fee Ti alloys have beenestablished. Besides the lattice constant, the Curie temperatureT C (123 K) and the saturation moment at 4.2 K have beenreported for stoichiometric L2 1-ordered Fe 2AlTi aged at900 C for 36 h [109]. For Fe3Al x Ti ( x 0e 11 at.%) theCurie temperature and the magnetization at 77 and 298 K have been determined [81]. Magnetisation and susceptibilitymeasurements in the temperature range 4.2 e 300 K have alsobeen carried out for FeAl (1 x )Ti x alloys [88]. Yamada et al.[110] studied the magnetic properties of the C14 Laves phaseon alloys (Fe 1 x Al x )2Ti with x 0.5 and found that antiferro-

magnetism in Fe 2Ti is suppressed by Al substitution and re-placed by ferromagnetism. The transition temperature fromferro- to para-magnetism goes through a maximum at about150 K for x 0.25.

The electrical resistance has been measured between 600and 800 C in order to determine the solid solubility of Fein a Ti [111] and in some Fe 3Al-based alloys [84]. Mossbauerspectra for alloys along Fe 3 x Ti x Al can be found in Refs.[85,87] .

9. Thermochemical, atomistic and diffusion data

There are few data available for the ternary Ale

Fee

Ti sys-tem. Bros et al. [112] measured the heat content of variousTie Ale X alloys, amongst which an alloy with 9.5 at. % Aland 2.1 at. % Fe. They used drop calorimetry up to 600 Cand compared their results with measurements on pure Ti.Thermodynamic properties of alloys of aluminium with ironand titanium have been studied by Gerashchenko and Temno-gorova [113] .

There are also only few data available on diffusion withinthe Ale Fee Ti system. Only recently diffusion data for tracerdiffusion of 59 Fe within TiAl single crystals [114] and poly-crystalline Fe 3Al-based alloys [115] have been reported. Fi-nally, it is worth to mention that stress induced phasetransformation in two-phase a Ti bTi alloys has been studiedby Koike et al. [116] .

10. Modelling

Databases for CALPHAD-type calculations have beenmade available for each of the corners of the Al e Fee Ti sys-tem. The concerted European COST action 507 has led to thedevelopment of a database for aluminium alloys [117] forwhich assessments were however limited to the descriptionof binary and some ternary systems, but not the Al e Fee Tione. In fact, no CALPHAD-type description of the whole ter-

nary Ale Fee Ti system is available at present. As a matter of

fact, the only tentative CALPHAD-type assessment of the Al eFee Ti system is due to Dew-Hughes and Kaufman [118] whocalculated the 1000 C isotherm. The extension of both theLaves phase and the B2-ordered phase FeTi in the ternary sys-tem was unexpected for these authors, who assumed that Alshould have substituted Ti in both phases, and this led them

to perform new experiments which just conrmed previousdata. Their conclusion, which appears still valid, is that furthertheoretical studies should attack the problem.

Schon [119] carried out a CVM calculation of the metasta-ble bcc Al e Fee Ti phase diagram, i.e. he calculated thetransitions between the A2, B2 and DO 3 structures at 800,900 and 1000 C. If only these three phases are consideredto be stable, then the B2 structure forms one single-phase eldinvading a large part of the composition triangle at 1000 C.The calculations also show that two DO 3 elds exist, one athigh Ti content and the other in the Fe-rich corner, which en-large with decreasing temperature. Emphasis was then put onthe various features of the rst- and second-order transitions inthe Fe-rich corner in comparison with experimental results ob-tained independently [78].

Ishikawa et al. [120] evaluated the interchange energy be-tween Al and Ti for the next nearest neighbour site when Feis in the nearest neighbour site in L2 1-ordered Fe 2AlTi.They also calculated the phase equilibria between FeAl (B2)and Fe2AlTi (L2 1) in the temperature range 1000 e 1300 Cand compared them with experimental results. The criticalordering boundary between a Ti (disordered A2) and bTi(ordered B2) at 1000 C has been obtained using the Bragg eWilliams e Gorsky approximation [71]. The same method hasalso been used to calculate the inuence of Fe on ordering

in D019 -type Ti 3Al [121] . Partition coefcients for Fe forphase equilibria involving the phases a Ti, bTi, Ti 3Al andTiAl have been evaluated [72]. Details of the kinetics duringordering and disordering in bcc Fe e Ale Ti alloys have beendescribed by use of the micro-master equation method [122] .The phase equilibria in the same composition range havebeen calculated by use of CVM by Ohnuma et al. [78].

Acknowledgement

This work has been carried out within the frame of COSTAction 535, Thermodynamics of alloyed aluminides(THALU).

References

[1] Zwicker U, Breme J, Nigge K. Mikrochim Acta Suppl 1985;11:333.[2] Zwicker U. Z Metallkd 1986;77:714.[3] Gartner S, Holze J. Aluminium 1989;65:1057.[4] Gorzel A, Palm M, Sauthoff G. Z Metallkd 1999;90:64.[5] Palm M. Intermetallics 2005;13:1286.[6] Diehm RS, Mikkola DE. Mater Res Soc Symp Proc 1987;81:329.[7] Prakash U, Muraleedharan K, Buckley RA, Jones H, Shenton PA.

J Mater Sci 1996;31:1569.[8] Nishino Y, Asano S, Ogawa T. Mater Sci Eng 1997;A234 e A236:271.[9] Sundar RS, Kutty TRG, Sastry DH. Intermetallics 2000;8:427.

ARTICLE IN PRESS


+ MODEL

7/28/2019 alfe2

12/13

[10] Zhu SM, Sakamoto K, Tamura M, Iwasaki K. Scripta Mater2000;42:905.

[11] Prakash U, Sauthoff G. Intermetallics 2001;9:107.[12] Zhu SM, Sakamoto K, Tamura M, Iwasaki K. Mater Trans JIM

2001;42:484.[13] Stein F, Schneider A, Frommeyer G. Intermetallics 2003;11:71.[14] Palm M, Sauthoff G. Intermetallics 2004;12:1345.[15] Pak HR, Wayman CM, Favrow LH, Cooper CV, Pak JSL. Mater Res

Soc Symp Proc 1991;186:357.[16] Morris DG, Gu nter S. Acta Metall Mater 1992;40:3065.[17] Winnicka MB, Varin RA. Metall Mater Trans 1993;A24:935.[18] Machon L, Sauthoff G. Intermetallics 1996;4:469.[19] Palm M, Gorzel A, Letzig D, Sauthoff G. In: Nathal MV, Darolia R,

Liu CT, Martin PL, Miracle DB, Wagner R, et al, editors. Structuralintermetallics 1997. Warrendale: TMS; 1997. p. 885.

[20] Ducher R. Contribution a` letude du diagramme de phase ternaireFee Tie Al et des equilibres Fe e TiAl. Thesis. Toulouse: INP; 2003.

[21] Brady MP, Smialek JL, Brindley WJ. Oxidation-resistant Ti e AleFe alloy for diffusion barrier coatings. US Patent no. 5 776 617;1998.

[22] Raghavan V. Phase diagrams of ternary iron alloys, part 1. Metals Park:ASM Int; 1987. 9.

[23] Raghavan V. J Phase Equilib 1993;14:618.[24] Raghavan V. J Phase Equilib 2002;23:367.[25] Kumar KS. Int Mater Rev 1990;36:293.[26] Ghosh G. Aluminium e iron e titanium. In: Petzow G, Effenberg G,

editors. Ternary alloys, vol. 5. Weinheim: VCH; 1992. p. 456.[27] Villars P, Prince A, Okamoto H. Handbook of ternary alloy phase dia-

grams, vol. 3. Metals Park: ASM Int; 1995. p. 3628.[28] Murray JL. Phase diagrams of binary titanium alloys. Metals Park:

ASM; 1987. p. 99.[29] Murray JL. Phase diagrams of binary titanium alloys. Metals Park:

ASM; 1987. p. 12.[30] Okamoto H. J Phase Equilib 1993;14:120.[31] Okamoto H. J Phase Equilib 2000;21:311.[32] Raghavan V. J Phase Equilib 2005;26:171.[33] Schuster JC, Palm M. J Phase Equilib Diff, accepted for publication.

[34] Kaufman L, Bernstein H. Computer calculation of phase diagrams.In: Refractory materials, vol. 4. New York: Academic Press; 1970.p. 183.

[35] Kaufman L, Nesor H. In: Jaffee RI, Burte HM, editors. Titanium sci-ence and technology, vol. 2. New York: Plenum Press; 1973. p. 773.

[36] Kaufman L, Nesor H. Calphad 1978;2:325.[37] Gros JP, Sundman B, Ansara I. Calphad 1988;22:1587.[38] Murray JL. Metall Trans 1988;A19:243.[39] Mishurda JC, Lin JC, Chang YA, Perepezko JH. Mater Res Soc Symp

Proc 1989;133:57.[40] Kattner UR, Lin J-C, Chang YA. Metall Trans 1992;A23:2081.[41] Oehring M, Klassen T, Bormann R. J Mater Res 1993;8:2819.[42] Saunders N. In: Ansara I, editor. COST507: thermochemical database

for light metal alloys. Brussels: ECSC e EECe EAEC; 1995. p. 52.[43] Saunders N. In: Blenkinsop P, Evans WJ, Flower HM, editors. Titanium 95:

science and technology, vol. 3. London: The Inst. Materials; 1996. p. 2167.[44] Zhang F, Chen SL, Chang YA, Kattner UR. Intermetallics 1997;5:471.

[45] Saunders N. COST507: thermochemical database for light metal alloys.In: Ansara I, editor. Brussels: ECSC e EECe EAEC; 1995. p. 89.

[46] Ohnuma I, Fujita Y, Mitsui H, Ishikawa K, Kainuma R, Ishida K. ActaMater 2000;48:3113.

[47] Danilenko VM, Srorchak-Veyduk AN, Yagodkin VV, Bulanova MV,Tretyachenko LA. In: Abstract of the sixth International school-confer-ence phase diagrams in Mater Sci PDMS-VI, Kiev; 2001. p. 233.

[48] Kattner UR. In: Massalski TB, editor. Binary alloy phase diagrams,vol. 1. Materials Park: ASM Int; 1990. p. 147.

[49] Inden G, Pepperhoff W. Z Metallkd 1990;81:770.[50] Ikeda O, Ohnuma I, Kainuma R, Ishida K. Intermetallics 2001;9:755.[51] Bichara C, Inden G. Scripta Metall Mater 1991;25:2607.

[52] Scho n CG, Inden G, Eleno LTF. Z Metallkd 2004;95:459.

[53] Mekhrabov AO, Masharov SI, Babaev ZM, Kazymov MM. J Turk Phys1994;18:444.

[54] Bamola RK, Seigle LL. Metall Trans 1989;A20:2561.[55] Stein F, Sauthoff G, Palm M. Z Metallkd 2004;96:469.[56] Lendvai AJ. Mater Sci Lett 1986;5:1219.[57] Aliravci CA, Pekgu leryuz O. Calphad 1998;22:147.[58] Pashechko M, Vasyliv C. Z Metallkd 1997;88:484.[59] Ellner M, Mayer J. Scripta Metall Mater 1992;26:501.[60] Burkhardt U, Grin Y, Ellner M, Peters K. Acta Crystallogr 1994;

B50:313.[61] Ellner M, Burkhardt U. J Alloys Compd 1993;198:91.[62] Freiburg C, Grushko B. J Alloys Compd 1994;210:149.[63] Burkhardt U, Ellner M, Grin Y, Ipser H, Richter K, Schuster JC, et al.

In: Final report international workshop on aluminium e transition metalphase diagrams. Graz, Austria; 1997. p. 54.

[64] Palm M. J Alloys Compd 1997;252:192.[65] Eumann M., Phasengleichgewichte und mechanisches Verhalten im

ternaren Legierungssystem Fe e Ale Mo. Thesis. Aachen: Shaker-Verlag;2002.

[66] Ducher R, Stein F, Viguier B, Palm M, Lacaze J. Z Metallkd2003;94:396.

[67] Palm M, Inden G, Thomas N. J Phase Equilib 1995;16:209.[68] Markiv VY, Burnashova VV, Ryabov VP. Akad Nauk Ukr SSR. Metal-

loz 1973;46:103.[69] Levin L, Tokar A, Talianker M, Evangelista E. Intermetallics

1999;7:1317.[70] Dew-Hughes D. Metall Trans 1980;A11:1219.[71] Kainuma R, Ohnuma I, Ishikawa K, Ishida K. Intermetallics

2000;8:869.[72] Kainuma R, Fujita Y, Mitsui H, Ohnuma I, Ishida K. Intermetallics

2000;8:855.[73] Hao YL, Yang R, Cui YY, Li D. Acta Mater 2000;48:1313.[74] Hao YL, Xu DS, Cui YY, Yang R, Li D. Acta Mater 1999;47:1129.[75] Anderson IM. Acta Mater 1997;45:3897.[76] Mekhrabov AO, Akdeniz MV. Acta Mater 1999;47:2067.[77] Banerjee R, Amancherla S, Banerjee S, Fraser HL. Acta Mater

2002;50:633.

[78] Ohnuma I, Scho n CG, Kainuma R, Inden G, Ishida K. Acta Mater1998;46:2083.[79] Mendiratta MG, Ehlers SK, Lipsitt HA. Metall Trans 1987;A18:509.[80] Fortnum RT, Mikkola DE. Mater Sci Eng 1987;91:223.[81] Sellers CH, Hyde TA, OBrien TK, Wright RN. J Phys Chem Solids

1994;55:505.[82] Anthony L, Fultz B. Acta Metall Mater 1995;43:3885.[83] Wright RN, Wright JK, Hyde TA, Sellers CM, Mishima Y. In:

Giamei AF, Inoue K, editors. Mechanical properties and phase trans-formations of multi-phase intermetallic alloys. Warrendale: TMS;1995. p. 49.

[84] Nishino Y, Kumada C, Asano S. Scripta Mater 1997;36:461.[85] Athanassiadis G, Le Caer G, Foct J, Rimlinger L. Phys Status Solidi

1977;A40:425.[86] Sun ZQ, Yang WY, Shen LZ, Huang YD, Zhang BS, Yang JL. Mater Sci

Eng 1998;A258:69.[87] Azez KA, Al-Omari IA, Shobaki J, Hasan MK, Al-Zoubi GM,Hamdeh HH. Physica B 2002;321:178.

[88] Okpalugo DE, Booth JG, Faunce CA. J Phys F Met Phys 1985;15:681.[89] Yanson TI, Manyako NB, Bodak OI, Cerny R, Gladyshevskii RE,

Yvon K. J Alloys Compd 1995;219:135.[90] Grytsiv A, Rogl P, Giester G, Pomjakushin V. Intermetallics

2005;13:497.[91] Rogl P, Grytsiv A. Presented at: Workshop Laves phases IV. Du sseldorf;

2005.[92] Mazdiyasni S, Miracle DB, Dimiduk DM, Mendiratta MG,

Subramanian PR. Scripta Metall 1989;23:327.[93] Kogachi M, Minamigawa S, Nakahigashi K. Scripta Metall Mater

1992;27:407.[94] Winnicka MB, Varin RA. Metall Trans 1992;A23:2963.

[95] Seibold A. Z Metallkd 1981;72:712.

ARTICLE IN PRESS


+ MODEL

7/28/2019 alfe2

13/13

[96] Ducher R, Viguier B, Lacaze J. Scripta Mater 2002;47:307.[97] Levin L, Tokar A, Berner A. J Mater Sci 2000;35:3923.[98] Tokar A, Berner A, Levin L. Mater Sci Eng 2001;A308:13.[99] Tokar A, Levin L, Katsman A, Ginzburg A, Berner A, Fein A, et al.

Mater Sci Eng, A351; 2003. p. 56.[100] Palm M, Zhang LC, Stein F, Sauthoff G. Intermetallics 2002;10:523.[101] Hata S, Shiraishi K, Itakura M, Kuwano N, Nakano T, Umakoshi Y.

J Electron Microsc 2004;53:537.[102] Kainuma R, Sato J, Ohnuma I, Ishida K. Intermetallics 2005;13:784.[103] Nishimura H, Matsumoto E. Nippon Kinzoku Gakkaishi 1940;4:339.[104] Nwobu A, Maeda T, Flower HM, West DRF. User aspects of phase

diagrams. Institute of Metals; 1991. p. 102.[105] Yang TY, Goo E. Metall Mater Trans 1995;A26:1029.[106] Palm M, Inden G. In: Nathal MV, Darolia R, Liu CT, Martin PL,

Miracle DB, Wagner R, et al, editors. Structural intermetallics 1997.Warrendale: TMS; 1997. p. 73.

[107] Mabuchi H, Nagayama H, Tsuda H, Matsui T, Morii K. Mater TransJIM 2000;41:733.

[108] Braun J, Ellner M. Metall Mater Trans 2001;A32:1037.[109] Buschow KHJ, van Engen PG. J Magn Magn Mater 1981;25:90.[110] Yamada Y, Kuroda K, Sakata A, Murakami T. J Magn Magn Mater

1998;177 e 181:1397.[111] Luzhnikov LP, Novikova VM, Mareev AP. Met Sci Heat Treat

1963;1963:78.

[112] Bros H, Michel ML, Castanet R. J Therm Anal 1994;41:7.[113] Gerashchenko SL, Temnogorova NV. Tekhnologiya i oborudovanie

proizvodstva tsvetnik i chernik metallov i splavov. Zaporozhye Indus-trial Institute; 1991. p. 4.

[114] Terashita N, Nose Y, Ikeda T, Nakajima H, Inui H, Yamaguchi M. MaterLett 2003;57:3357.

[115] Golovin IS, Divinski SV, Cizek J, Prochazka I, Stein F. Acta Mater2005;53:2581.

[116] Koike J, Shimoyama Y, Ohnuma I, Okamura T, Kainuma R, Ishida K,et al. Acta Mater 2000;48:2059.

[117] Ansara I, Dinsdale AT, Rand MH. Thermochemical database for lightmetal alloys. Luxembourg: European Commission; 1998.

[118] Dew-Hughes D, Kaufman L. Calphad 1979;3:175.[119] Scho n CG. Thermodynamics of multicomponent systems with chemical

and magnetic interactions. Thesis. Universita t Dortmund; 1998.[120] Ishikawa K, Kainuma R, Ohnuma I, Aoki K, Ishida K. Acta Mater

2002;50:2233.[121] Kalyuzhnyi VV, Matysina ZA, Milyan MI. Sov Phys J 1987;30:242.[122] Ni J, Ashino T, Iwata S. Acta Mater 2000;48:3193.[123] Kornilov II , Pylaeva EN, Volkova MA. Russ J Inorg Chem 1958;3:169.[124] Volkova MA, Kornilov II . Russ Metall 1970;1971:134.[125] Volkova MA, Kornilov II . Russ Metall 1971;1971:137.[126] Kainuma R, Urushiyama K, Ishikawa K, Jia CC, Ohnuma I, Ishida K.

Mater Sci Eng 1997;A239 e 240:235.

ARTICLE IN PRESS


+ MODEL

Documents

alfe2