Iron aluminides: present status and future prospects

(1)

Iron aluminides: present status and future prospects

N.S. Stoloff

Materials Science and Engineering Department, Rensselaer Polytechnic Institute, Troy, NY 121280-3590, USA

Abstract

This paper constitutes a broad survey of the physical, mechanical and corrosion properties of Fe Al alloys, as well as a₃ review of principal processing methods. This class of alloys, once thought to be inherently brittle, is shown to possess moderate ductility, provided that mechanical testing is carried out in an inert environment. Methods to improve mechanical properties by alloying and microstructural control are described. The influence of alloying elements on corrosion and stress corrosion resistance and weldability also is reviewed. Q 1998 Elsevier Science S.A. All rights reserved.

Keywords: Fracture; Impact; Iron aluminides; Corrosion welding

1. Introduction and historical review

The iron aluminides Fe Al and FeAl have been₃ among the most widely studied intermetallics because of their low cost, low density, good wear resistance, ease of fabrication and resistance to oxidation and corrosion. These advantages have led to the identifi- cation of several potential uses, including heating elements, furnace fixtures, heat-exchanger piping, sin- tered porous gas-metal filters, automobile and other industrial valve components, catalytic converter subs- trates and components for molten salt applications w x1,2 . In addition, Fe Al is one of the few structural₃ intermetallics that can be disordered with respect toŽ DO -type order by appropriate elevated temperature₃ .

w x

heat treatment, as shown in Fig. 1 3 . This pheno- menon is both an advantage for scientific studies ofŽ the influence of ordering on mechanical behavior. and a disadvantage, owing to the degradation of creep and tensile strengths upon disordering. Further exac- erbating this disadvantage is the low temperature, approx. 5508C, at which disordering occurs. This criti- cal temperature, T , becomes the limiting tempera-_c ture for structural applications, although it can be raised substantially by alloying with silicon or molyb- denum, among other solutes. At present, some of the major issues that continue to delay commercial viabil- ity include low ductility and impact resistance at low temperatures and inadequate creep resistance at elevated temperatures.

The physical properties, mechanical behavior and corrosion resistance of Fe Al alloys have been exten-₃

w x

sively reviewed in recent conferences 4,5 as well as w x

in other volumes on intermetallics 6,7 . Accordingly, only a brief summary of the early work on these alloys will be presented here. Most of the review will be devoted to the current status of research and development efforts on iron aluminides as well as a discus- sion of prospects for commercial applications.

The excellent corrosion resistance of Fe]Al alloys w x

was first recognized in the 1930s 8 , but detailed studies of mechanical behavior commenced with the work of Cahn and his co-workers in the late 1950s

w x

and early 1960s 9]11 . This work included reports of the creep resistance of Fe Al above and below the₃

w x

critical ordering temperature, T_c 9 , as well as a description of the effect of long-range order on yield-

w x w x

ing 10 and mechanical twinning 11 . A discontinuity in the slope of plots of creep rate vs. inverse tempera- ture near T was attributed to a change in the activa-_c

w x tion energy for diffusion at this temperature 9 . Later work by Stoloff and Davies dealt with the influence of

w x

long-range order on yielding 12 and strain hardening w x13 of Fe Al. A study of compositional effects on₃ w x yielding of Fe]Al alloys was reported by Sainfort 14 . Other noteworthy early work by Justusson and

w x w x

Zackay 15 and by Kayser 16 on fracture behavior of Fe Al should be cited. These studies showed that₃ as aluminum is added to iron ductility drops sharply,

Ž .

P I I S 0 9 2 1 - 5 0 9 3 9 8 0 0 9 0 9 - 5

(2)

Fig. 1. Variation of DO₃ order determined from variations in w x X-ray as a function of annealing temperature before quenching 3 .

especially as the Fe Al phase becomes stable. The₃ early work indicated that the low ductility of ‘disordered’ Fe Al was further reduced slightly by order-₃

w x

ing 15 . It was established also that a peak in yield stress at an intermediate degree of long-range order occurs at room temperature in quenched Fe Al as₃

w x

well as at elevated temperature 13,15 . This work, together with studies of the FeCo]V intermetallic, led to the suggestion that the flow stress peak in many intermetallics is due to a transition from motion of single dislocations to superlattice dislocations as the degree of long-range order increases from zero to one w12,16 . This theory has since been supplanted byx

more complex arguments concerning dislocation core structures, but still serves as a readily understandable model of the role of superlattice dislocations in plastic flow behavior of intermetallics. Furthermore, the electron microscopic studies of Marcinkowski and

w x

Brown 17 at approximately the same time provided information about slip systems, dislocation configura- tions and antiphase boundaries in these alloys.

Efforts to develop iron aluminides as structural materials were carried out under Air Force sponsor-

w x

ship at Pratt and Whitney Aircraft 18 and Marko w x

Materials 19 , among others. The superior corrosion resistance of Fe Al in aqueous solutions was well₃ known, but it was recognized that utilization of alloys based on Fe Al would be limited by the relatively low₃ temperature at which long-range order is lost approx.Ž 5508C . Accordingly, efforts were made to improve. high-temperature strength by adding hard particles, but these efforts ultimately were terminated without

w x any resulting applications 18 . 2. Phase relationships

The currently accepted Fe]Al phase diagram, in w x the vicinity of 25 at.% Al, appears in Fig. 2 20 . The Fe Al region ranges from approx. 223 ]30 at.% Al, but two-phase regions exist on either side. Two-phase alphaqB2 and alphaqDO regions noted in the3

diagram must be accounted for when determining heat treatment schedules for these alloys. In the two-phase regions, age hardening behavior which can substantially alter mechanical properties has been

w x w x

reported 1,21,22 . For example, Morris 1 has noted

w x Fig. 2. Phase diagram in vicinity of Fe]25 at.% Al 20 .

(3)

that in the temperature range of approx. 300]5508C and for alloys containing 20]25 at.% Al, it is possible to retain a stable two-phase alphaqDO ordered3

microstructure similar to the gamma]gamma prime microstructures in nickel-base superalloys. The temperature range of stability of this two-phase mixture can be extended by adding Si in the range of 1]3%.

In the range 23]27 at.% Al, the B2 structure can only be retained at low temperatures by quenching from above T , but at higher Al contents B2 replaces DO_c ₃ as the stable phase. The wide solubility range of B2 FeAl allows it to exist in the partially ordered condi- tion above T on either side of stoichiometry._c

3. Dislocations and antiphase boundaries

Plastic deformation in dilute, disordered Fe]Al al-

² :

loys is accomplished by the motion of unit a 111_o dislocations, as in pure iron, and wavy slip indicates that cross slip readily occurs among planes containing the slip vector. In hyperstoichiometric Fe Al, super-₃ lattice dislocations are observed, of the fourfold type in the DO condition, and paired in the B2 condition,₃

² :

but always of 111 type. At the stoichiometric composition uncoupled dislocations often are seen. At high temperatures deformation has been thought to

² :

be controlled by the glide of perfect 100 disloca-

w x w x

tions 23 ; however, Kad and Horton 61 report that

² :

only 111 dislocations are present in FeAl deformed at 925 K and in Fe Al deformed at 1075 K. Composi-₃ tional effects do not appear to account for the dis- crepancy, especially for the FeAl alloy, and texture measurements are not in agreement with the activa-

² :

tion of 100 slip systems. Therefore further work is required to unambiguously identify operative dislocations at high temperatures.

Thermally produced antiphase domains are found in both the B2 and DO₃ variants of Fe Al. Their₃ boundaries do not lie on preferred planes, resulting in a wavy, isotropic appearance that is easily imaged by transmission electron microscopy. There is no evidence that these boundaries inhibit dislocation motion. However, unit dislocations trying to move through the fully ordered DO₃ lattice would leave behind a trail of antiphase boundary, the energy for which has to be supplied by the applied forces.

4. Processing

Iron aluminides are readily prepared in small quantities either by melting and casting or by powder processing. Alloys can be melted by a variety of tech-

Ž .

niques, including air induction melting AIM , vac-

Ž .

uum induction melting VIM and vacuum arc remelt-

Ž .

ing VAR . Arc melting, vacuum induction melting and drop casting have been utilized to prepare high-

purity test materials. However, for commercialization to be practical, less expensive methods are needed.

Also, prevention of hydrogen gas uptake in the molten alloy is important to produce ingots free of voids.

Argon gas blown through the melt effectively elimi- w x

nates the porosity 4 . Another approach, carried out in India, is to air induction melt, followed by electros-

Ž . w x

lag remelting ESR 24 . Clean, defect-free ingots were obtained from remelting even of porous induction melted ingots. Improved microstructures and mechanical properties were obtained with cast ESR ingots containing 0.074]0.14 wt.% C. Cast Fe Al ingots3

are best reduced by hot working, typically at temperatures of 1000]12008C; warm working is then carried

w x

out between 6508C and 8008C 4 . In laboratory experiments, cast material tends to have much lower ductility than wrought samples of the same composition, due in large part to the coarse grain size and weak grain boundaries of the castings; as a result, such castings cannot be cold worked or used in the as-cast condition for structural applications.

The exothermic reaction between aluminum and iron can be utilized in both melting and powder consolidation. Although the exotherm is not as great as for the nickel aluminides, low ignition temperatures permit melting by the Exo-Melt^TM process, see

w x

Fig. 3 25 , or by powder processing via the reactive sintering self-propagating high-temperature synthe-Ž

. w x

sis 26]28 approach. Utilizing the exothermic reaction lowers costs and, in the case of melting, offers greater safety, shorter melt times and improved process control. Iron aluminide powders can be used to form near net shape parts by hot isostatic press-Ž ing , or can be used to produce spray coatings. Powders. are typically prepared by gas atomization, utilizing nitrogen, argon or helium. Spherical particles usually result, with oxygen contents approximately the same as in the melt in purged systems with pure carrier gas w x27 . Nitrogen gas atomized alloy FAS see Table 1Ž . prepared by reactive sintering has been shown to have slightly higher yield and tensile strengths at temperatures to 8008C than cast product, but much higher creep resistance; elongations of the material prepared by the two methods were similar.

Hot pressing of elemental powders resulted in the formation of single phase Fe Al with 98.2% of theo-₃ retical density. Other successful powder techniques included hot extrusion and a combination of mechani-

w x

cal alloying and reactive sintering 26 . Excellent mechanical properties were attributed to uniformity of microstructures and fineness of grain sizes. Other powder techniques that have been reported include mechanical alloying of prealloyed, atomized powders

w x

with small amounts of Y O₂ ₃ 29 , injection molding of w x

Fe Al with short Al O fibers 30 and thermal spray-₃ ₂ ₃ ing of elemental Fe and Al powders followed by

(4)

Table 1

a w x

Compositions of iron aluminides chosen at the Oak Ridge National Laboratory for commercialization 2 Ž .

Element Alloy %

b c d e

FAS FAL FA-129 FAPY

Weight Atomic Weight Atomic Weight Atomic Weight Atomic

Al 15.9 28.08 15.9 28.03 15.9 28.08 8.46 16.12

Cr 2.20 2.02 5.5 5.03 5.5 5.04 5.50 5.44

B 0.01 0.04 0.01 0.04 ] ] ] ]

Zr ] ] 0.15 0.08 ] ] 0.20 0.11

Nb ] ] ] ] 1.0 0.51 ] ]

C ] ] ] ] 0.05 0.20 0.03 0.13

Mo ] ] ] ] ] ] 2.00 1.07

Y ] ] ] ] ] ] 0.10 0.06

Fe 81.89 69.86 78.44 66.81 77.55 66.17 63.71 77.07

aFAS, FAL, and FA-129 are the Fe Al-base alloys. FAPY is a disordered3 a alloy.

bSulfidation-resistant alloy.

cHigh room-temperature tensile ductility.

dHigh-temperature strength with good room-temperature ductility.

eVery high room-temperature ductility.

w x

annealing to produce the intermetallic 26 . High-energy ball milling also has been used to form nanocrys-

w x

talline iron aluminides 31 . Control of composition in all of these methods is very important, although Fe Al₃ is less sensitive than FeAl in this regard due to the absence of constitutional vacancies in the former.

5. Alloy development

As aluminum is added to iron, strength increases

and ductility decreases; ductility changes are particularly pronounced as the ordering range, above 16

w x

at.%, is reached 23 . It is now known that sensitivity to moisture is the primary cause of low ductility in Fe Al alloys at room temperature, although contami-₃ nation by impurities such as carbon is a contributing factor. Chromium is the most effective solute to com- bat environmental embrittlement, and as a conse- quence a series of Fe]Al]Cr alloys has been devel- oped at the Oak Ridge National Laboratory, see

w x Fig. 3. Furnace-loading sequence to take advantage of heat of formationof Fe Al during the melting of iron-aluminide alloys 25 .₃

(5)

Table 1. Optimum ductility is produced at the w x

28Al]5Cr level 23 . Zirconium additions also improve ductility, but the effect depends upon whether

w x

carbon is present 32,33 . The mechanism for solute improvements is not yet known, although interference with the moisture dissociation reaction at an Al-rich surface is a likely candidate.

Alloys containing Si, Ta, Ce, Zr, Mo, Hf or Nb have been shown to increase high-temperature creep and tensile strength of Fe Al, usually at the expense of₃

w x w x

room-temperature ductility 22 . Chen et al. 34 , sug- gest that Mo is the most effective of these solutes,

Ž .

possibly because of its effects on T_c raised by Mo , the increased APB energy and reduced diffusivity.

The low solubilities of these solutes, as well as their tendency to form intermetallic compounds or borides or carbides, allows them to be used for precipitation hardening and for grain refinement. Quantities of these solutes up to 10 at.%, together with very small additions of boron or carbon, can be used to improve both tensile and creep strength with little adverse

w x

effect on low-temperature ductility 23 . Carbide strengthening is particularly effective in Fe Al, but₃ melting and subsequent processing should insure that carbides will precipitate from the liquid, thereby producing refined cast structures and subsequent fine- grained wrought product. Heat treatment schedules are particularly important in optimizing creep properties of Fe Al alloys. For example, an alloy containing₃ small additions of Nb, Mo, Zr, B and C showed much longer creep lives at 5938C under a stress of 207 MPa when heat treated at 11508C rather than at 7508C w x35 . This effect was attributed to the dissolution of coarse carbides and subsequent precipitation of fine particles on matrix dislocations.

6. Mechanical properties 6.1. Yielding

The yield stress of binary Fe]Al alloys increases as Al content increases, to a peak at the stoichiometric composition; strength then decreases until B2 FeAl forms. The strength of Fe Al is related to the pres-₃ ence of long-range DO -type order and the occur-₃ rence of two-phase alphaqDO regions as Al con-3

w x tent exceeds 16 at.%. Although McKamey 23 at- tributes strengthening to the interaction of superlattice dislocations with thermally produced antiphase boundaries and boundaries produced by superpartial glide dislocations, this seems unlikely, based on ear-

w x

lier work of Stoloff and Davies 16 on Cu Au.₃ Quenching of Fe Al alloys to produce partial B2-type₃ order results in a small decrease in flow stress.

As in the case of many other intermetallics, the flow stress of Fe Al alloys displays a peak with in-₃

Fig. 4. Temperature dependence of the critical resolved shear stresst of Fe]28 at.% AlqCr single crystals deformed in uniaxial_o

w x

compression.B: No Cr, orientation 123 ; I: 6 at.% Cr, orienta-

w x w x w x

tion 123 ; 0:6 at.% Cr, orientation 001 36 .

creasing test temperature, in the range 450]5508C, depending upon composition; the peak occurs at w x slightly lower temperatures at higher Al contents 23 . Although most data have been obtained for polycrys-

w x

tals, Kral et al. 36 have reported data for Fe]28 at.% Al and Fe]28Al]6Cr single crystals. As shown in Fig. 4, the critical resolved shear stress decreases in region A, just above room temperature, and then rises to a peak near 8508K. Cr produces some softening at temperatures below the peak. No evidence of a rela- tion between the state of long-range DO -type order₃ and the peak was observed. Early theories of the flow stress peak associated it with the loss of DO -type₃ order as the critical ordering temperature, T , was_c

w x

approached. However, Schroer et al. 37 showed that the peak could still be seen in DO -based alloys₃ modified to retain order at the temperature of the

w x

peak. Morris 1 has recently analyzed the various theories that have been advanced to explain the peak:

4 4

1. extensive cross slip between 110 and 112 ; 2. APB relaxation, leading to dislocation drag on

²111 dislocations;:

² : ² : ² :

3. transition from 111 to 110 and 100 dislocations with increasing temperature;

4. superdislocation climb locking; and 5. vacancy hardening.

Strain rate change tests as well as temperature jump experiments carried out in order to deduce the operative mechanism were inconclusive, leading to the conclusion that several mechanisms may control deformation over a range of temperature, strain rate

w x and alloy composition 1 .

w x

Stoloff and Davies 16 showed that there is a maximum in yield stress with the degree of quenched in long-range order at room temperature upon quenching from just below T . This peak seemed to_c be related to that obtained at temperature, in that

(6)

both occurred at an intermediate degree of long-range order. The recent investigations cited by Morris w x1 above do not seem to have considered the im- plications of these observations. More recent

w x work has been reported by Lu et al. 38 for a Fe]28Al]5Cr]0.3B]0.003% Mg alloy. Room-temperature tensile strength rises to a maximum for specimens quenched in oil from 600]6508C subsequent to

w x

a stress relief heat treatment, see Fig. 5 38 . Cooling rate has a significant effect, with furnace-cooled material displaying lower tensile strength and ductility than oil-quenched samples, probably because of the different ordered states existing in the two cases.

6.2. Tensile strength and ductility

Since iron aluminides are often considered to be potential competitors to stainless steels for structural and corrosion-resistant applications, it is interesting to compare tensile properties of several wrought Fe Al alloys with types 310 and 422 stainless steels,₃

w x

see Fig. 6a,b 2 . Note that yield and tensile strengths

Ž .

of FAS, FAL and FA-129 see Table 1 are intermediate between those of the stainless steels, and a shal- low peak in yield strength is exhibited by all of the

w x

aluminides. Similarly, Fig. 6c 2 shows that the room- temperature ductility of the aluminides is intermediate to those of the stainless steels, but at higher temperatures a transition to ductile behavior occurs for the former. As a result, the total elongation of the aluminides exceeds those of the stainless alloys at temperatures above 5508C. A more extensive discus- sion of low-temperature ductility appears in the sec- tion on environmental embrittlement.

6.3. Impact properties

Limited studies of impact behavior of stress-relieved Fe Al alloys has revealed disappointingly low₃ absorbed energies and a high ductile]brittle transi-

Ž .

tion temperature 3088C , as shown in Fig. 7 for the w x

FA-129 alloy 38 . Furthermore, the upper shelf energy is very low, only 37 J, while the lower shelf

w x w x

energy is only 5 J 39 . Liu et al. 40 have reported impact data for Fe]25% Al and Fe]25Al]20Mn.

They too found very low-impact energies at the lower shelf for the binary alloy, but the ternary alloy was much tougher. Peaks in total impact energy vs. temperature were noted for both alloys near 6008C, con- siderably higher than reported above for FA-129.

However, above the peak the impact energy dropped sharply for the ternary alloy and hardly at all for the binary alloy. The addition of Mn introduced an ordered L1 -type phase, and decreased the DO order-₂ ₃ ing energy of the alpha phase. These factors appear to be responsible for both increased toughness and a transition from cleavage to dimpled rupture with the addition of the Mn.

6.4. Fracture toughness

Most fracture studies have been conducted with tensile specimens. However, there are now limited

Ž .

fracture toughness data including J integral data w x

available for binary Fe]28 at.% Al 40 . As in the case of tensile experiments, environmental effects are significant, with oxygen atmospheres providing higher w x fracture resistance than vacuum or air, see Fig. 8 41 . Note that toughness in oxygen is double that in air for

w x Fig. 5. Room temperature tensile properties vs. heat-treatment temperatures for oil quenched specimens 38 .

(7)

Fig. 6. Comparison of average tensile properties of wrought Fe Al-based alloys with that of types 422 and 310 stainless steels: a yield₃ Ž .

Ž . Ž . w x

strength; b ultimate tensile strength; and c total elongation 2 .

Fe]28% Al, with an even greater disparity noted for

Ž .

Fe]35% Al B2 structure . However, there is no dif- ference in toughness in any environment between the DO and partially ordered B2 structures, unlike the₃ case of tensile elongation, which is higher in the B2 condition. In general, the toughness of Fe Al is higher₃ than that of other aluminides such as TiAl and NiAl.

6.5. Creep and stress rupture

The creep resistance of binary Fe Al alloys is rela-₃ tively poor, largely due to the open B2 and DO₃ crystal structures. Therefore a major aim of alloy

Ž .

development efforts see above has been to improve creep resistance while not reducing low-temperature ductility.

The stress sensitivity of creep rate and stress exponent have been determined for several Fe Al alloys,₃

w x

as shown in Table 2 42 . Both parameters are stress dependent, although the average stress exponent is in the range 3]7. Similarly, the most common value for activation energy is 300]350 kJrmol. Note that several different mechanisms of creep are observed.

Binary and more complex Fe Al solid solution al-₃ loys display lower creep rupture resistance than many

competing structural alloys such as types 310 and 422 w x

stainless steels, see Fig. 9 2 . Note that in spite of comparable tensile strengths, the creep rupture lives of the steels are much superior. Fortunately, it is possible to improve rupture lives of Fe]28 at.% Al by addition of Ti, Nb, Zr, B and especially Mo, as shown

w x

in Fig. 10 43 . The combination of 2% Moq0.1Zrq

Ž .

0.2B alloy FA-114 is particularly effective, with that alloy showing the highest creep resistance of several alloys tested. It was suggested that fine ZrC particles pin dislocations, thereby improving both strength and

w x

creep resistance 43 . Another factor that can markedly affect creep resistance is heat treatment.

Fig. 11 shows that the highest creep-rupture life for

Ž .

alloy FA-180 see Table 1 occurs at a heat treatment w x

temperature of 11508C for 1 h 43 . Fine carbides precipitate during this treatment. The creep strength of this alloy compares favorably with that of 316 and 403 stainless steels. Heat treatments at 11508C for more than 4 h result in shorter lives, possibly due to carbide coarsening. Rupture lives are increased further by quenching from 11508C into either oil or water.

Chromium, which is usually present for resistance to environmental embrittlement, slightly lowers rup-

(8)

w x Fig. 7. Absorbed energy vs. temperature for full-size T-L specimens of iron]aluminide alloy FA-129 39 .

Fig. 8. Fracture toughness of iron aluminides in different environ- w x

ments 41 .

ture lives of B2 Fe Al at 6003 8C and 200 MPa, while w x increasing the tendency for intergranular fracture 34 . The same investigation revealed that Cr lowers the APB energy, resulting in larger separations of twofold dislocations and a greater tendency for motion of uncoupled dislocations. The result is larger elongations and increased stress concentrations at grain boundaries, thereby accounting for more intergranular failure.

w x

Chen et al. 34 report that grain shape has little effect on rupture life of Fe]28Al]2Cr, but grain size is an important variable. Fig. 12 shows that rupture life increases to a maximum as grain size increases

w x

from 70 mm to 372 mm 34 . Alloys with Mo, Cr, Zr

and Nb also show increased lives, up to 2000 h, at larger grain sizes.

6.6. Superplasticity

Superplasticity has been observed in coarse grained w x FeAl and Fe Al alloys by Lin and co-workers 43 .₃ Maximum elongations of 620% have been noted for Fe]28 at.% Al]2Ti at 8508C under an initial strain rate of 1.26=10^y3rs. At this temperature strain hardening is slight, while at 700 and 7508C distinct

w x

strain hardening is noted, see Fig. 13 44 . The following criteria for superplasticity have been identi- fied:

Ž .

1. high strain rate sensitivity m)0.3 ; 2. low, strain independent flow stresses;

3. temperature at least 0.5T ; and_m 4. high ductility.

The unusual feature of superplasticity in the iron aluminides is the large grain size at which it occurs Ž60]100 mm for Fe Al and 500]600 mm for FeAl ;3 . these grain sizes may be compared to the 1]5 mm usually necessary for superplasticity in conventional metals and alloys.

7. Environmental resistance 7.1. En_¨ironmental embrittlement

The results of early work on the iron aluminides had suggested that they were intrinsically brittle at

w x

low temperatures 13]16 . However, it was shown by

w x

Liu and co-workers 45,46 that when water vapor and

(9)

Table 2

w x Summary of creep parameters in Fe Al and its alloys 42₃

Ž . Ž . Ž .

Alloy T K Q_e kJrmol n Mechanism comments Ref

w x

Fe]19.4Al 773]873 305 4.6]6 for Diffusion controlled 9

Fe]27.8Al 823]888 276 different Al

Higher temperature 418 ] Controlled by state of order

Ž w x

Fe]15]20Al )773 260 to 305 increases ] Diffusion controlled 57

with increasing Al.

-773 s dependent ] Motion of jogged screw

dislocations

Ž .

Fe]28Al 898 347 3.5 lows Viscous glide

Ž .

7.7 highs Climb

Ž . w x

Fe]28Al]2Mo 923 335 1.4 lows Diffusional flow 58

Ž .

6.8 highs Climb

Ž .

Fe]28Al]1Nb] 923 335 1.8 lows Diffusional flow

Ž .

0.013Zr 19.0 highs Dispersion strengthening

w x

FA-180 866 627 7.9 Precipitation strengthening 59

w x

Fe]28Al 873]948 ] 3.4 Viscous glide 60

Fe]26Al]0.1C 873]948 305 3.0 Viscous glide

753]813 403 6.2 ?

Fe]28Al]2Cr 873]948 325 3.7 Viscous glide

Fe]28Al]2Cr]0.04B 873]948 304 3.7 Viscous glide

Fe]28Al]4Mn 873]948 302 2.6 Viscous glide

Fig. 9. Comparison of creep rupture properties of wrougt Fe Al-₃ w x

based alloys with types 422 and 310 stainless steels 23 .

other sources of hydrogen are excluded from the environment, ductilities in excess of 10% were achiev- able at room temperature. The deleterious effects of hydrogen-containing environments are seen in other aluminides, especially Ni Al, TiAl and Ti Al, as well₃ ₃

w x

as in Ni Si and other intermetallics 47 . Even in₃ moist air, the ductility of Fe Al can be increased by₃ adding Cr or Zr andror by producing a stress-relieved

w x but not recrystallized microstructure 23 . Also, quenching from above T to form the partially-ordered_c

Fig. 10. Creep-rupture data showing the effect of additions of Zr w x and heat teatment on the creep strength of Fe]28Al]2Mo 43 .

B2 structure is beneficial. Consequently, adequate ductility for most structural applications can be achieved in the alloys listed in Table 1. Interestingly, fracture occurs by transgranular cleavage, independent of the environment, except for alloys containing Cr, which exhibit mixed intergranular and transgranular crack segments. Also, when heat treatment of Fe]28Al]5Cr subsequent to a stress relief anneal is

w x

applied, Lu et al. report 38 intergranular fracture after slow cooling from the annealing temperature.

The environmental effects on fracture in tension described above are replicated under cyclic loading.

Fatigue crack growth rates are much lower in vacuum or oxygen than in air or in hydrogen gas, as shown, for

(10)

w x Fig. 11. Creep-rupture life vs. heat treatment for 1-h anneals of alloy FA-180. Specimens were air cooled following heat treatment 43 .

Fig. 12. Effects of annealing temperature on the creep properties of Fe]28 at.% Al]2 at.% Cr alloy at a stress of 200 MPa and 6008C w x34 .

Ž .

example Fe]15]9 wt.% Al]5.5Cr]1.0Nb]0.05C in w x

Fig. 14 44 . Fractography reveals transgranular cleav- w x

age in air or hydrogen gas 47 , but microvoid coales- cence is observed for tests carried out in oxygen.

Similar effects of gaseous environments on growth rates have been observed on several other Fe Al₃

w x alloys 48 .

7.2. Aqueous corrosion and stress corrosion

The aqueous corrosion behavior of Fe Al alloys₃ containing 28 at.% Al heat treated to produce B2Ž order in various corrosive media have been reported.

w x

by Buchanan and co-workers 49,50 , with their results reproduced in Table 3. High corrosion rates are noted

Fig. 13. True stress]true strain for Fe]28Al]2Ti showing super- w x

plastic behavior at 7508C and 8508C 44 .

in acid- and sulfur-bearing environments, especially relative to the reference material, 304L SS. In 1 molrl NaOH, on the other hand, all four materials are passivated at the corrosion potential, E_corr, and remain passivated to high potentials; this is indicative of good overall corrosion resistance. Low average corrosion rates are seen also in chloride solutions, but localized corrosion initiates in FA-84 and FA-129 alloys within 24 h. The addition of 1]2% Mo and 4]6% Cr is beneficial in delaying initiation, as shown

w x

in Table 4 50 . Tests of these alloys under crevice corrosion conditions mild acidŽ ]chloride solutions for 18 h showed inferior behavior relative to 304L SS,. indicating that crevice corrosion can occur even in Fe Al containing Cr and Mo.₃

Embrittlement caused by hydrogen released from water vapor in air has been documented in many studies of tensile stress]strain behavior. Similarly, stress corrosion of Fe Al alloys in aqueous solutions₃ has been linked to the evolution of hydrogen, result-

(11)

w x Ž . Ž . Fig. 14. Fatigue crack growth of Fe Al alloy FA3 ]129 258C 48 : a B2; b DO .3

w x

ing in embrittlement. Kasul and Heldt 51 have shown that ductility of Fe]24.6% Al is decreased at cathodic potentials in both acid and basic environments, see Fig. 15, exactly as expected from test conditions favoring release of hydrogen. Similarly, slow strain experiments carried out in air in a mild acid]chloride solution show decreasing ductility with more negative potentials as well as at the highest anodic potential,

w x

see Fig. 16 50 . The latter observation seems to be connected to the formation of corrosion pits during exposure, leading to accelerated hydrogen production in the presence of a stress concentrator. The maximum ductilities are observed near the free corrosion potential and are comparable to the ductility in air.

Even the disordered Fe]16 at.% Al alloy exhibits stress corrosion behavior as well as very high fatigueŽ w x.

crack growth rates in hydrogen and in moist air 52 , in sharp contrast to the high ductility and dimpled fracture mode resulting from tensile tests. In summary, low strain rates, cathodic potentials and conditions favoring pitting are detrimental to the stress corrosion behavior of both ordered and disordered Fe]Al alloys containing 16]28 at.% Al.

7.3. Oxidation and hot corrosion

The high-temperature corrosion behavior of Fe Al₃ alloys has recently been reviewed by Natesan and

w x

Tortorelli 53 . The formation of Al O films provides₂ ₃ oxidation resistance; the minimum Al content to form Al O is 162 3 ]18 at.% and these films can form even in low pressures of oxygen. Small additions of Cr can reduce the minimum Al content somewhat. The mechanical integrity of the films and their reformation characteristics are critical to adequate behavior at long exposures. When Fe Al is exposed to sulfur-₃ bearing atmospheres, weight gain is adversely affected by the presence of large quantities of Cr, as shown in

w x

Fig. 17 53 . Note that the binary Fe]12 wt.% Al alloy as well as Fe Al alloy FA-61 shows no weight gain at₃ 8758C. Apart from Cr, the only other solute which has been studied in detail is Zr, which appears to improve

w x oxide adherence at elevated temperatures 54 . 8. Welding

Considerable attention has been devoted to the weldability of Fe Al alloys because of concern about₃ the problem of hydrogen embrittlement in the pres-

w x

ence of water vapor 55,56 . All traces of the latter must be excluded from the weld surface in order to avoid delayed cold cracking. Reducing grain size has proven to be an effective means of improving weld ductility and increasing resistance to hydrogen embrittlement. Another method to control cold cracking

Table 3

w x Corrosion rates by the polarization-resistance method for Fe Al-based iron aluminides in acidic, basic and sulfur-compound solutions 50₃

Ž .

Solutions Average penetration rates mmrday

Materials

FA-84 FA-129 FAL-Mo 304L SS

1 molrl HCI 430 52 14 9.7

0.5 molrl H SO₂ ₄ 400 72 120 0.014

1 molrl HNO₃ 160 3.5 1.4 0.007

1 molrl NaOH 0.042 0.021 0.056 0.007

0.1 molrl Na S O₂ ₂ ₃ 7.8 9.9 2.6 0.004

0.1 molrl Na S O₂ ₄ ₆ 21 6.5 6.5 0.020

(12)

Table 4

Localized-corrosion initiation times for Fe Al-based₃ iron w x

aluminides in a mild acid-chloride solution 50

Iron aluminides Localized-corrosion

initiation times Ždays.

Fe]28Ala 1

Fe]28Al plus:

2Cra 2

Ž .b

2Cr]0.05B FA-84 1

4Cra 2

Ž .

5Cr]0.2C]0.5Nb FA-129 1

6Cra 27

1Moa 2

2Cr]1Moa 41

4Cr]0.5Moa 59

a c

4Cr]1Mo )122

a c

4Cr]2Mo )122

b c

Ž .

5Cr]1Mo]0.04B]0.08Zr FAL-Mo )259

304L Stainless steel )259c

aDO heat treatment.₃

bB2 heat treatment.

cTests terminated.

is to preheat at 3508C or higher or post-heat at 7508C, in order to relieve stresses and drive off hydrogen w x55 . These temperatures are higher than those previ-

w x ously suggested by McKamey 23 .

Hot cracking of iron aluminides is not as pervasive as cold cracking, and is highly dependent on composition. Zirconium, boron and TiB are detrimental to₂ weldability, while niobium, chromium and small amounts of carbon are beneficial. Threshold cracking stresses are similar to those of some austenitic stain-

w x less steels and Ni Al 23 .₃

Fig. 15. Ductility of an Fe]24.6% Al alloy under conditions of

Ž .

control potential and pH electrochemical potentials are vs. SCE w x51 .

Fig. 16. Slow-strain-rate ductility vs. electrochemicals potential for w x

FA-129 iron aluminide 50 .

9. Summary

This review has dealt with the highlights of recent research on the processing, microstructure, mechanical properties and environmental resistance of Fe Al₃ alloys. While much progress has been made in solving the twin problems of poor low-temperature ductility and inadequate high-temperature creep resistance, Fe Al alloys appear still to be a few years away from₃ widespread applications. In part this situation stems from the contradictory effects of chromium on

Ž .

strength lowered and low-temperature ductility Žraised . As others solutes are added for improved. creep strength, ductility may be expected to decrease

Fig. 17. Weight change data for Fe]Cr, Fe]Cr]Ni, Fe]Al and Fe]Cr]Al alloys, and Fe aluminide Fe Al tested in OrS environ-3

ment with pO2 4.1=10^y18 and pS2s9.4=10^y7 atm at 8758C w x53 .

(13)

again; in any case, ductility can be expected to approach zero when atomic hydrogen from the environment or within the alloy is mobile. Therefore care must be taken in processing, cutting or joining to protect alloys from moisture or other sources of hydrogen. While coatings may alleviate the problem, their utilization entails extra expense, and if the coat- ing is damaged, protection is lost. In spite of these difficulties, Fe Al alloys continue to offer advantages₃ over other structural materials, as outlined in the introduction, and alloy development continues. To date, most of such development in the United States has been conducted at the Oak Ridge National Laboratory. It is desirable to expand alloy development efforts to other laboratories here and abroad in order to continue and accelerate the momentum of ORNL’s efforts.

Acknowledgements

The author is grateful to the US Department of Energy, Fossil Energy AR& TD Materials Program for financial support under subcontract No. 19X-SF- 521C with Lockheed Martin Energy Research Corp.

References

w x1 D.G. Morris, in: S.C. Deevi, V.K. Sikka, P.J. Maziasz, R.W.

Ž .

Cahn Eds. , International Symposium on Nickel and Iron Aluminides, ASM International, Materials Park, OH, 1997, p.

73.

w x2 V.K. Sikka, in: S.C. Deevi, V.K. Sikka, P.J. Maziasz, R.W.

Ž .

Cahn Eds. , International Symposium on Nickel and Iron Aluminides, ASM International, Materials Park, OH, 1997, p.

361.

w x3 D.G. Morris, in: J.H. Schneibel, M.H. Crimp Eds. , Process-Ž . ing, Properties and Applications of Iron Aluminides, TMS, Warrendale, PA, 1994, p. 117.

w x4 V.K. Sikka, in: J.H. Schneibel, M.H. Crimp Eds. , Processing,Ž . Properties and Applications of Iron Aluminides, TMS, War- rendale, PA, 1994, p. 3.

w x5 S.C. Deevi, V.K. Sikka, P.J. Maziasz, R.W. Cahn ŽEds. ,. International Symposium on Nickel and Iron Aluminides, ASM International, Materials Park, OH, 1997.

w x6 N.S. Stoloff, V.K. Sikka Eds. , Physical Metallurgy and Pro-Ž . cessing of Intermetallic Compounds, Chapman and Hall, New York, 1996.

w x7 J.H. Westbrook, R.L. Fleischer Eds. , Intermetallic Com-Ž . pounds, John Wiley, New York, 1996.

w x8 N. Ziegler, Trans. AIME 100 1932 267.Ž .

w x9 A. Lawley, J.A. Coll, R.W. Cahn, Trans. Met. Soc. AIME 218 Ž1960 166..

w x10 A. Lawley, E.A. Vidoz, R.W. Cahn, Acta Metall. 9 1961 287.Ž . w x11 R.W. Cahn, J.A. Coll, Acta Metall. 9 1961 138.Ž .

w x12 N.S. Stoloff, R.G. Davies, Acta Metall. 12 1964 473.Ž . w x13 R.G. Davies, N.S. Stoloff, Acta Metall. 11 1963 1187.Ž .

w x14 G. Sainfort, P. Mouturat, P. Pein, J. Petit, G. Cabane, M.

Ž .

Salesse, Mem. Sci. Rev. Met. 60 1963 125.

w x15 F.X. Kayser, Ford Motor Co., WADC Tech. Rept. 57]298 Žpart 1 , 1957..

w x16 N.S. Stoloff, R.G. Davies, Prog. Mat. Sci. 13 1966 3.Ž . w x17 M.J. Marcinkowski, N. Brown, Acta Metall. 9 1964 764.Ž . w x18 R.G. Bordeau, Development of Iron Aluminides, AFWAL-

TR-4009. Wright Patterson AFB, OH, 1987.

w x19 R. Ray, V. Panchanathan, S. Isserow, J. Metals 35 1983 30.Ž . w x20 P.R. Swann, W.R. Duff, R.M. Fisher, Metall. Trans. 3 1972Ž .

409.

w x21 H. Inoue, in: C.C. Koch, C.T. Liu, N.S. Stoloff Eds. , HighŽ . Temperature Ordered Intermetallic Alloys, MRS Sym. Proc., 39, MRS, Pittsburgh, 1985, p. 255.

w x22 G.E. Fuchs, N.S. Stoloff, Acta Metall. 36 1988 1381.Ž . w x23 C.G. McKamey, in: N.S. Stoloff, V.K. Sikka Eds. , PhysicalŽ .

Metallurgy and Processing of Intermetallic Compounds, Chapman and Hall, New York, 1996, p. 351.

w x24 R.G. Baligidad, U. Parakash, A. Radhakrishna, V.R. Rao, P.K. Rao, N.B. Ballal, in: S.C. Deevi, V.K. Sikka, P.J. Mazi-

Ž .

asz, R.W. Cahn Eds. , International Symposium on Nickel and Iron Aluminides, AMS International, Materials Park, OH, 1997, p. 177.

w x25 V.K. Sikka, in: N.S. Stoloff, V.K. Sikka Eds. , Physical Metal-Ž . lurgy and Processing of Intermetallic Compounds, Chapman and Hall, New York, 1996, p. 561.

w x26 S.C. Deevi, V.K. Sikka, in: S.C. Deevi, V.K. Sikka, P.J. Mazi-

Ž .

asz, R.W. Cahn Eds. , International Symposium on Nickel and Iron Aluminides, ASM International, Materials Park, OH, 1997, p. 283.

w x27 J.K. Wright, R.N. Wright, in: J.H. Schneibel, M.H. Crimp ŽEds. , Processing, Properties and Applications of Iron. Aluminides, TMS, Warrendale, PA, 1994, p. 47.

w x28 K.G. Shaw, J.H. Reinshagen, in: S.C. Deevi, V.K. Sikka, P.J.

Ž .

Maziasz, R.W. Cahn Eds. , International Symposium on Nickel and Iron Aluminides, ASM International, Materials Park, OH, 1997, p. 301.

w x29 M.N. Srinivasan, V.K. Sikka, in: J.H. Schneibel, M.H. Crimp ŽEds. , Processing, Properties and Applications of Iron. Aluminides, TMS, Warrendale, PA, 1994, p. 69.

w x30 J. Scott, MS Thesis, Rensselaer Polytechnic Inst., 1992.

w x31 R. Frattini, J. Schiffine, G. Scipione, E. Bonett, S. Enzo, in:

Ž .

J.H. Schneibel, M.H. Crimp Eds. , Processing, Properties and Applications of Iron Aluminides, TMS, Warrendale, PA, 1994, p. 79.

w x32 V.K. Sikka, Oak Ridge National Lab., unpublished.

w x33 D.A. Alven, Ph.D. Thesis, Rensselaer Polytechnic Inst. 1997.

w x34 G. Chen, Y. Huang, W. Yang, Z. Sun, in: J.H. Schneibel,

Ž .

M.H. Crimp Eds. , Processing, Properties and Applications of Iron Aluminides, TMS, Warrendale, PA, 1994, p. 131.

w x35 C.G. McKamey, P.J. Mazsiaz, in: J.H. Schneibel, M.H. Crimp ŽEds. , Processing, Properties and Applications of Iron. Aluminides, TMS, Warrendale, PA, 1994, p. 147.

w x36 F. Kral, P. Schwander, G. Kostorz, in: S.C. Deevi, V.K. Sikka,

Ž .

P.J. Maziasz, R.W. Cahn Eds. , International Symposium on Nickel and Iron Aluminides, ASM International, Materials Park, OH, 1997, p. 337.

w x37 W. Schroer, H. Mecking, C. Hartig, in: O. Izumi ŽEd. ,. Proceedings of the International Symposium On In-

Ž .

termetallics JIMIS 6 , Japan Inst. of Metals, 1991, p. 567.

w x38 Y.H. Lu, Z.Q. Xing, H. Zhang, Scripta Metall. 33 1995 609.Ž . w x39 B.G. Gieseke, D.J. Alexander, V.K. Sikka, R.H. Baldwin,

Ž .

Scripta Metall. 29 1993 129.

w x40 J.N. Liu et al., in: C.C. Koch, C.T. Liu, N.S. Stoloff, A.

Ž .

Wanner Eds. , High Temperature Ordered Intermetallic Al- loys VII, MRS Symp. Proc. 460, MRS, Pittsburgh, 1997, p.

343.

(14)

w x41 S.H. Ko, R. Gnanamoorthy, S. Hanada, Mat. Sci. Eng. A222 Ž1997 133..

w x42 D.H. Sastry, R.S. Sundar, in: S.C. Deevi, V.K. Sikka, P.J.

Ž .

w x43 N.S. Stoloff, D.A. Alven, C.G. McKamey, in: S.C. Deevi, V.K.

Ž .

Sikka, P.J. Maziasz, R.W Cahn Eds. , International Sympo- sium on Nickel and Iron Aluminides, ASM International, Materials Park, OH, 1997, p. 65.

w x44 D. Lin, in: S.C. Deevi, V.K. Sikka, P.J. Maziasz, R.W. Cahn ŽEds. , International Symposium on Nickel and Iron. Aluminides, ASM International, Materilas Park, OH, 1997, p.

187.

w x45 C.T. Liu, C.G. McKamey, E.H. Lee, Scripta Metall. 24 1990Ž . w x46 385.C.T. Liu, in: I. Baker, R. Darolia, J.D. Whittenberger, M.H.

Ž .

Yoo Eds, , High Temperature Ordered Intermetallic Alloys V, MRS. Symp. Proc. 288, MRS. Warrendale, PA, 1993, p. 3.

w x47 N.S. Stoloff, in: N.S. Stoloff, V.K. Sikka ŽEds. , Physical. Metallurgy and Processing of Intermetallic Compounds, Chapman and Hall, New York, 1996, p. 479.

w x48 N.S. Stoloff, Iron Steel Inst. J. Japan 37 1997 1197.Ž . w x49 R.A. Buchanan, J.G. Kim, Proceedings of the Sixth Annual

Conference On Fossil Energy Materials, Oak Ridge Natl.

Lab., 1992, p. 315.

w x50 R.S. Buchanan, J.G. Kim, R.E. Ricker, L.A. Heldt, in: N.S.

Ž .

Stoloff, V.K. Sikka Eds. , Physical Metallurgy and Processing of Intermetallic Compounds, Chapman and Hall, 1996, p.

517.

w x51 D.B. Kasul, L.S. Heldt, in: R.H. Jones, R.E. Ricker Eds. ,Ž .

Environmental Effects on Advanced Materials, TMS, War- rendale, PA., 1991, p. 67.

w x52 A. Castagna, N.S. Stoloff, in: J.A. Horton, I. Baker, S. Hanada,

Ž .

R.D. Noebe, D.S. Schwartz Eds. , High Temperature Or- dered Intermetallics V, MRS Symp. Proc. 364 2 , MRS, Pitts-Ž . burgh, 1995, p. 1171.

w x53 K. Natesan, P.F. Tortorelli, in: S.C. Deevi, V.K. Sikka, P.J.

Ž .

w x54 P.F. Tortorelli, B.A. Pint, I.G. Wright, Proceedings of the 11^th Annual Conference On Fossil Energy M aterials, ORNLrFMP-97r7, 1997, p. 235.

w x55 M.L. Santella, in: S.C. Deevi, V.K. Sikka, P.J. Maziasz, R.W.

Ž .

Cahn Eds. , ASM International Symposium on Nickel and Iron Aluminides, ASM International, Materials Park, OH, 1997, p. 321.

w x56 S.A. David, M.L. Santella, in: N.S. Stoloff, V.K. Sikka Eds. ,Ž . Physical Metallurgy and Processing of Intermetallic Com- pounds, Chapman and Hall, New York, 1996, p. 655.

w x57 R.G. Davies, Trans AIME 227 1963 22.Ž .

w x58 C.G. McKamey, P.J. Maziasz, J.W. Jones, J. Mat. Res. 7 Ž1992 2089..

w x59 V.K. Sikka, B.G. Gieseke, R.H. Baldwin, in: K. Natesan, D.J.

Ž .

Tillack Eds. , Heat Resistant Materials, ASM, Materials Park, OH, 1991, p. 363.

w x60 R.S. Sundar, D.H. Sastry, Indian Inst. of Science, Bangalore, unpublished, 1996.

w x61 B.K. Kad, J.A. Horton, Mater. Sci. Eng., 1997;A239-240:118.