444

Tjong

Werkstoffe und Korrosion37,444-447 (1986)

Stress corrosion cracking of the austenitic Fe-Al-Mn

alloy in chloride environment
SpannungsriBkorrosion von austenitischen FeAlMn-Legierungen in Chloridlosungen
S. C. Tjong*

The SCC behavior of Fe-8.7A1-29.7Mn-1.04C alloy in a 20%

sodium-chloride solution (100"C) has been investigated by the slow
strain rate method. The results showed that the alloy was susceptible
to SCC only at the free corrosion potential and at the applied strain
rate of 2.8 x lO-'/s. Application of the cathodic potential during the
siow strain rate test resulted in the loss of ductility in the alloy which
was attributed to the hydrogen induced cracking.

1 Introduction
The austenitic Fe-Al-Mn alloys have recently attracted considerable interest of the researchers for use as substitutes for
the conventional Cr-Ni stainless steels. The A1 and Mn replace
Cr and Ni, respectively in the Cr-Ni stainless steels. The FeAl-Mn alloys are cheaper than Cr-Ni stainless steels since Cr
and Ni are considered to be expensive and strategic materials
in some countries. These new alloys have a low density and
good oxidation resistance due to their A1 content. The representative Fe-Al-Mn alloys that have been investigated by the
researchers are Fe-lOA1-34.5Mn-0.76C and Fe-8A1-30Mn1.5Si-1.OC [l-21. The alloying elements C and Mn are known
to stabilize the austenitic structure which gives rise to good
high temperature mechanical properties. On the other hand,
A1 is a femte stabilizer and therefore its amounts must be
confined to certain limits. Addition of Si up to 2% increases
the oxidation resistance and mechanical strength of the alloys,
but decreases the austenite stability [2-31.
According to the Fe-Al-Mn phase diagrams [4], a brittle
p-Mn phase could be present in the alloys containing more
than 35% Mn. Wang and Beck [5] obtained a similar finding
and reported that the formation of @-Mn phase could be
avoided in alloys containing 20 to 30% Mn. The optimal composition of Fe-Al-Mn alloy having good mechanical properties
should contain about 8-10% A1 and 25-30% Mn [6].
The oxidation resistance of Fe-Al-Mn alloys has been
reported by S u e r et al. [7]. They showed that the alloys containing 1to 2% Si have good oxidation resistance at 850C due
to the formation of a continuous alumina layer. However, the
corrosion resistance of Fe-Al-Mn alloys in aqueous environments is little known. Wang and Beck [5] have compared the
polarization curves of Fe-1OAl-30Mn-Si alloy and 18Cr-9Ni-Ti
stainless steel in seawater. Their preliminary results showed
that the corrosion rate of Fe-1OAl-30Mn-Si alloy is somewhat
Institute of Materials Science & Engineering, National Sun Yat-Sen
University, Kaohsiung, Taiwan.

Das SpannungsriSkorrosionsverhalten der Legierung Fe-8, 7 A1-29,

7 Mn-l,04 C in 2O%iger Natriumchloridlosung von 100 "C ist mit
Hilfe der ,,slow strain rate"-Methode untersucht worden. Die Ergebnisse zeigen, daR die Legierung bei der Dehngeschwindigkeit 2 , 8 . W 7 /
s nur am freien Korrosionspotential anfallig fur SpannungsriSkorrosion ist. Im kathodischen Potentialbereich trat unter den Versuchsbedingungen ein Duktilitatsverlust auf, welcher der wasserstoffinduzierten RiBbildung zugeschrieben wurde

lower than that of 18Cr-9Ni-Ti stainless steel. On the basis of

the potentiodynarnic curves and the simple immersion test,
they concluded that the Fe-10A1-30Mn-Sialloy can be used as
a candidate material for seagoing ship propellers. However,
for the effective and safe use of the propellers, the stress corrosion cracking (SCC) resistance of the alloy is one of the most
important properties. Furthermore, in choosing from among
the candidates when replacing the Cr-Ni stainless steels which
are frequently used in high temperature aqueous environments, the SCC behavior of the Fe-Al-Mn alloys at high temperatures must also be considered. In this paper, a preliminary
investigation of the SCC behavior of Fe-Al-Mn alloy in 20%
sodium-chloride solution at 100"C has been carried out by the
slow strain technique. This method involves pulling the tension specimen at a slow rate of about or less than 10-'/s. The
occurrence of SCC on the specimen was identified by the
changes in the mechanical properties and fracture appearances. This method is effective in producing cracks in a short
period and a better data reproducibility than that of the constant load technique. Further details of the application of the
slow strain method to determine the SCC susceptibility have
been discussed elsewhere [Sl.

2 Experimental
The Fe-8.7A1-29.7Mn-l.UC alloy was prepared by melting
high purity iron, aluminium, and manganese in a laboratory
induction furnace. It was then cast into a steel mold. The cast
ingot was hotforged at 1200C into a billet. This billet was
homogenized at 1100C for 12 hours, air cooled and subsequently cold-rolled to 2 mm thick plates. The tensile specimens having a gauge length of 30 mm were machined from the
billet with the specimens axis parallel to the direction of rolling. After final machining, the tensile specimens were solution-treated at 1050"C for 2 hours and subsequently quenched
into oil upon removal from the furnace. The gauge length of
the specimen was abraded down to 600 grit, it was then
0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim. 1986

Werkstoffe und Korrosion 37, 444-447 (1986)

Stress corrosion cracking of the austenitic Fe-Al-Mn alloy

400r

Table. Mechanical properties of the Fe-8.7A1-29.7Mn-1.04Calloy

Tabelle. Mechanische Eigenschaften der untersuchten Legierung
Fe-8,7A1-29,7Mn-l,04C
alloy

0.2% proof stress tensile strength

(kg/mm2)
(kgimm2)

Fe-8.7AI-29.7Mn- 36.40
1.04c

69.80

445

elongation

(%I
52.92

degreased with acetone, and washed with distilled water. The

non-gauge part of the tensile specimen was covered with teflon
tape. Details of the mechanical properties of the solution-treated specimens are given in the Table.
The slow strain rate experiment was run with an Instron
universal testing machine (model 1125). The crosshead speeds
used in the test were from 0.5 mm/min to 0.00005 mm/min and
these correspond to relative strain rates of 2.8 x 10-4/s to
2.8 x 10%.
The electrochemical test cell used was a glass cylinder which
consisted of a platinum counterelectrode, a condenser, a mercury thermometer and a capillary which was connected to a
beaker in which a silver-silver chloride reference electrode was
placed. The test solution was made from reagent grade
sodium-chloride and distilled water. It was then heated to
100C by an electric resistance tape coiled around the glass
cylinder. An ECO potentiostat (model 553) was used to control the potential of the specimen.
After testing, the fracture surface of the specimen was ultrasonically cleaned in acetone in order to remove the corrosion
products and subsequently examined in the scanning electron
microscope. Percent elongation was determined from measurements of the final specimen length. It is difficult to measure the reduction of area due to the small size of the specimens employed.

3 Results and Discussion

Figure 1 shows the microstructure of the solution-treated
Fe-8.7A1-29.7Mn-l.MC alloy. As can be seen, the microstructure of the alloy investigated is fully austenitic. Fig. 2 shows

DISPLACEMENT. rnm

m
1
.
i

Fig. 2. Load-displacement curves of the Fe8.7Al-29.7Mn-1.04C steel

obtained with different strain rates in a 20% sodium-chloride solution
(100OC) at the open circuit potential
Abb. 2. Belastung-Verformung-Kurvendes Stahls Fe-8,7A1-29,7Mn1,MC bei verschiedenen Dehngeschwindigkeiten in 20%iger Natriumchloridlosung bei 100C ohne angelegtes Potential

some representatives of the load-displacement curves for the

Fe-8.7A1-29.7Mn-l.MC alloy tested in 20% sodium-chloride
solution at the open circuit potential (- 0.480 V, SHE) for
various strain rates. It is apparent from the curves that the
most severe ductility loss is observed at the employed strain
rate of 2.8 x lO-/s. The strain rate dependence of the elongation at fracture of Fe-8.7A1-29.7Mn-l.MC alloy is shown in
Fig. 3. It has been shown that the susceptibility of an alloy to
SCC depended on the applied strain rates [9]. At high strain
rates, the specimen experienced no SCC due to the mechanical
failure which occurred prior to the onset of SCC. However, at
an extremely low strain rate, SCC also does not occur because
the bare metal surface produced by the applied stress are
repassivated. The formation of new film during the repassivation process protects the underlying alloy from an active dissolution. Examination of the fracture surfaces of the specimens after slow strain rate tests reveal some important morphological differences. Fig. 4 shows the SEM fractograph of
the Fe-8.7A1-29.7-Mn-l.MC alloy which failed in 20% sodium
chloride solution at a strain rate of 2.8 x lO-/s. The surface

-I

lci8
:w l.GT]

Fig. 1. Photomicrograph of the structure of the austenitic Fe-8.7A129.7Mn-1.04C steel

Abb. 1. Mikroaufnahrne des Gefiiges des austenitischen Stahls Fe8,7A1-29,7Mn-l,04C

16

10%
STRAIN RATE , S

lo5

10

Fig. 3. Effect of strain rate on the elongation to fracture of the Fe8.7A1-29.7Mn-l.MC alloy in a 20% sodium-chloride solution (100C)
at the open circuit potential
Abb. 3. EinfluB der Dehngeschwindigkeit auf die Bruchdehnung des
Stahls Fe-8,7AI-29,7Mn-l,04C bei verschiedenen Dehngeschwindigkeiten in 20%iger Natriumchloridlosung bei 100C ohne angelegtes
Potential

446

Tjong

Fig. 4. Fractograph of the Fe-8.7A1-29.7Mn-1.04C alloy after testing

in a 20% sodium-chloride solution (l0OT) at open circuit potential.
and at the applied strain rate of 2.8 x lO-/s
Abb. 4. Bruchbild des Stahls Fe-8,7A1-29,7Mn-I ,04C nach Lagerung
in 2076iger Natriurnchloridliisung hei 100C ohne angelegtes Potential
bei der Dehngeschwindigkeit 2,8 . lW7s

Werkstoffe und Korrosion 37,444-447 (1986)

Fig. 6. SEM fractograph of the Fe-8.7A1-29.7Mn-l.WC alloy after

testing in a 20% sodium-chloride solution ( 100C) at the open circuit
potential, and the applied strain rate of 2.8 x
Abb. 6. REM-Bild der Bruchflache des Stahls Fe-8,7A1-29,7Mn1,04C nach Lagerung in 20%iger Natriumchloridlosung bei 100C
ohne angelegtes Potential bei der Dehngeschwindigkeit 2,8. W 5 / s
r

2 5

Fig. 5. SEM fractograph of the Fe-8.7A1-29.7Mn-1.04C alloy aftcr

testing in a 20% sodium-chloride solution (IOOC) at open circuit
potential, and the applied strain rate of 2.8 x IO-(/s
Abb. 5. REM-Bild der Bruchflache des Stahls Fe-8,7A1-29,7Mn1,04C nach Lagerung in 20%iger Natriumchloridlosung hei 100C
ohne angelegtes Potential bei der Dehngeschwindigkeit 2,8.10 (/s

appearance of the specimen showed a transgranular cleavage

mode fracture. One can see from this figure that the cleavage
facets were separated by numerous steps, and there are active
slip lines on the cleavage facets. It is well known that the
transgranular SCC occurs during the slow strain rate test by
the repetitive rupture of the passive films. The active slip system intersects the specimen surface which promotes rupture of
the passive films and thus causing an active dissolution at these
localized sites.
Figure 5 shows the fracture surface of the alloy strained at a
rate of 2.8 x 10% The surface showed transgranular quasicleavage, and cleavage fracture along with some secondary
cracks present. As the strain rate is increased to 2.8 X lO-/s,

.-

50

DISPLACEMENT

75
I

100

mm

Fig. 7. Effect of the applied potential on the load-displacement curves

for the Fe-8.7AI-29.7Mn-1.04C alloy in a 20% sodium-chloride solution
Abb. 7. EinfluB des angelegten Potentials auf die Belastung-Verformung-Kurven des Stahls Fe-8,7 Al-29,7 Mn-1.04 C in 20%iger Natriumchloridlosung

the amount of brittle cleavage decreased whereas the amount

of quasicleavage increased dramatically. There are numerous
dimples in the quasicleavage zone (Fig. 6). It is apparent from
Fig. 6 that the SCC crack started from the edges on both sides
of the specimen and the final rapid proceeded by the microvoid coalescence. Therefore, it is obvious that the fracture
surface morphology changed in parallel with mechanical properties. From the fractographs and mechanical properties discussed previously, a strain rate of 2.8 x lO-/s (a critical value)
is slow enough to prevent dimpled fracture, yet causes severe
SCC on the alloy investigated.
Figure 7 shows the load-displacement curve of Fe-8.7A129.7Mn-1.04C alloy in 20% sodium-chloride solution at the
applied potential of - 0.348 V (SHE) and a strain rate of 2.8
x lO-/s. The curve shows that polarizing the specimen anodi-

Werkstoffe und Korrosion 37,444-447 (1986)

cally resulted in substantial increase in the elongation to fracture and the ultimate fracture stress of the specimen. In other
words, anodic polarization reduces the chance for the occurrence of SCC.
On the other hand, as the applied potential was cathodic to
the open circuit potential, a decrease in ductility was observed
(Fig. 7). At the applied potential of - 0.528 V (SHE), some
bubbles of hydrogen can be seen evolving from the specimen
surface during straining. And the production of these hydrogen bubbles is more pronounced at the applied lower cathodic
potential (- 0.578 V, SHE). It is apparent that a decrease in
ductility was due to the hydrogen induced failure. The extensive plastic deformation during the slow strain test played a
key role in the hydrogen-assisted fracture. This is due to the
continuous production of bare metal surfaces at slip steps during straining. These bare surfaces enable sufficient amounts of
hydrogen to be adsorbed and transported into the specimen
during straining.

4 Summary
The susceptibility of Fe-8.7A1-29.7Mn-l.MC alloy to SCC
in a 20% sodium-chloride solution (100C) was tested by a
slow strain technique. The results showed that at the open
circuit potential, a strain of 2.8 X lO-/s was slow enough to
prevent dimpled fracture, and lead to a transgranular SCC.
Anodic polarization above the open circuit potential effectively increases the SCC resistance of the alloy investigated. In

Stress corrosion cracking of the austenitic Fe-Al-Mn alloy

447

contrast, cathodic polarization resulted in a decrease of the

ductility of the alloy. The embrittlement is attributed to the
evolving of hydrogen at the applied cathodic potentials, and
the adsorption and diffusion of hydrogen into the alloy during
the slow strain test.

5 Acknowledgements
The author would like to thank Prof. C. M. Wan of National
Tsing Hua University for kindly providing the alloy investigated. This work was supported by the National Science Council under grant number NSC 74-0201-E110-01.

References
1. J. L. Ham, R. E. Cairns: Product Engineering 29 (1958) 50.
2. D. J. Schmutz: Trans. ASM 52 (1960) 898.
3. J. Charles, A. Berghezan, A . Lutts, P. L. Dancokne: Metal Progress 123 (1981) 71.
4. V. G. Lidin: International Metals Reviews 28 (1983) 309.
5. R. Wang, F. H . Beck: Metal Progress 125 (1983) 72.
6. G. L. Kayak: Metal Science & Heat Treatment 2 (1969) 95.
7. 1. P. h e r , R. A . Rapp, J. P. Hirlh: Oxid. Met. 18 (1982) 285.
8. G. M . Ugiamky, 1. H . Payer: ASTM, STP 665, 1979.
9. C. D. Kim, B. E. Wilde in: ASTM STP 665 (Ed. G. M. Ugiumky
and 1. H. Payer), (1979) p. 97.

(Received: 06.08.1985)

W2413