See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/238368803

Lattice expansion of carbon-stabilized

expanded austenite

Article in Scripta Materialia · October 2010

DOI: 10.1016/j.scriptamat.2010.05.031

CITATIONS READS

21 91

3 authors, including:

Thomas Christiansen Marcel Somers

Technical University of Denmark Technical University of Denmark
61 PUBLICATIONS 832 CITATIONS 206 PUBLICATIONS 3,489 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Compound layer formation during nitriding and nitrocarburizing View project

Supermartensitic stainless steels for oil and gas applications View project

All content following this page was uploaded by Marcel Somers on 04 March 2014.

The user has requested enhancement of the downloaded file.

This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
 Author's personal copy

Available online at www.sciencedirect.com

Scripta Materialia 63 (2010) 761–763

www.elsevier.com/locate/scriptamat

Lattice expansion of carbon-stabilized expanded austenite

Thomas S. Hummelshøj, Thomas L. Christiansen and Marcel A.J. Somers*
Technical University of Denmark, Department of Mechanical Engineering, Kemitorvet b.204, DK 2800 Kgs. Lyngby, Denmark
Received 4 May 2010; revised 21 May 2010; accepted 21 May 2010
Available online 26 May 2010

The lattice parameter of expanded austenite was determined as a function of the content of interstitially dissolved carbon in
homogeneous, carburized thin stainless steel foils. For the first time this expansion of the face-centered cubic lattice is determined
on unstrained austenite. It is found that a linear relation holds between the lattice parameter, a, and the occupancy of the octahedral
interstices, yC. The results are compared to other values in the literature and discussed in terms of stressed and unstressed austenite.
Ó 2010 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

Keywords: X-ray diffraction (XRD); Stainless steels; Crystal structure; Surface alloying

Austenitic stainless steels are the most widely
used corrosion-resistant steels. Engineering these mate-
rials towards improvement of, particularly, the wear
resistance and, to a lesser extent, the corrosion perfor-
mance can be accomplished by interstitial dissolution
of relatively large amounts of carbon and/or nitrogen
into the surface by carburization, nitriding or nitrocar-
burizing at a temperature below which effectively no car-
bides or nitrides develop within the treatment time. This
brings about a supersaturated solid solution of carbon
and/or nitrogen in austenite, so-called expanded austen-
ite [1,2]. The lattice parameter of austenite, which de-
pends on the interstitial atom content, is essential in
the interpretation of expected residual stress in ex-
panded austenite grown into an austenitic substrate. A
detailed assessment of the stress-free lattice parameter
of nitrogen-stabilized expanded austenite was published
in Refs. [3,4]. Despite several attempts to determine the
dependence of the lattice parameter of (expanded) aus-
tenite on the carbon content [2,5], an accurate determi-
nation for the homogeneous stress-free state has so far
not been published. The major difficulty with the vast
majority of the expanded austenite samples hitherto
investigated with X-ray diffraction is that they suffer
from a composition gradient within the depth range
investigated with X-rays [2,5]. Consequently, the influ-
ence of this depth range and an associated compressive
stress gradient influence the values reported [6]. In the
* Corresponding author. Tel.: +45 45252250; e-mail: somers@mek.
dtu.dk
present work the influence of composition gradients
and composition-induced stress gradients are circum-
vented by the investigation of homogeneous samples
of uniform composition as obtained by carburizing
and subsequent homogenization of thin stainless steel
samples.
Thin foils of stainless steel AISI 316 of nominal com-
position 16–18 wt.% Cr, 10–14% Ni, 2–3% Mo, 2% Mn
with thickness 7.5 lm were used for carburizing. Gas-
eous carburizing was performed in a Netzsch 449C
simultaneous thermal analyzer (referred to as thermo-
balance in the sequel), which allows thermogravimetric
analysis (TGA) simultaneously with differential thermal
analysis (DTA). Gaseous carburizing was performed in
gas mixtures of acetylene, hydrogen and nitrogen (nitro-
gen was led through the measurement compartment of
the thermobalance to protect this part from reaction
with the other gases) [7]. The flows were adjusted with
Brooks electronic mass flow controllers to 0.2 or 50 ml
C2H2 min1, 50 ml H2 min1 and 5 ml N2 min1.
For recrystallization and austenitization the stainless
steel foils were heated in pure H2 to 1353 K at a heating
rate of 30 K min1 and upon reaching this temperature
immediately cooled. During austenitization, deforma-
tion-induced martensite, which had formed during roll-
ing of the foil, was transformed into austenite. The
austenitization was conducted in a separate, for this pur-
pose, dedicated furnace in order to make larger batches.
Subsequently the foils were mounted in the thermobal-
ance for carburizing, heated to a carburizing tempera-
ture of 693 K in pure hydrogen and nitrogen and
reached the carburizing temperature within approxi-

1359-6462/$ - see front matter Ó 2010 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.scriptamat.2010.05.031
 Author's personal copy

762 T. S. Hummelshøj et al. / Scripta Materialia 63 (2010) 761–763

mately 20 min. Upon reaching the carburizing tempera- 3.70

ture, the gas composition was changed to the mixture of
C2H2, H2 and N2. In order to obtain different carbon

Lattice parameter, a (Å)

a= 3.5965 + (0.6029± 0.0189) yC
contents, and to follow the development of expanded
austenite, 10 carburizing experiments were performed
on thin foils, yielding samples with different average car- 3.65
bon contents in the range 0.5–6 wt.% C. Subsequently,
the gas composition was changed to pure Ar followed
by immediate cooling. In order to correct for buoyancy
effects, an identical temperature–time–gas composition
program was run with empty crucibles. In order to 3.60
establish the relation between dissolved carbon and lat-
tice parameter, six of the as-carburized foils were subse- 0.00 0.05 0.10 0.15
quently treated in pure Ar for 5 h at a temperature of Interstitial content, yC
773 K to distribute the absorbed carbon atoms over
Figure 2. Lattice parameter, a, of expanded austenite as a function of
the thickness of the foil. The homogenization time was the number of atoms per metal atom (yC) in cC.
estimated with the model in Ref. [8], using a diffusion
coefficient of 4.1  1016 m2 s1 [9]. The thermobalance
allowed continuous monitoring of the sample weight, carbon atoms, i.e., the development of cC. As verified
ensuring that no decarburization occurred during the by the asymmetric broadening and splitting of the cC
homogenization. 2 2 0 peaks in Figure 1a, a pronounced carbon concen-
X-ray diffraction was used to characterize the as-pre- tration profile exists in the as-carburized foils, while
pared and carburized material. The instrument used was the center is still free of carbon, as reflected by the per-
a Bruker AXS D8 X-ray diffractometer, equipped with a sistence of the original austenite reflection. X-ray diffrac-
Cr anode and a set of Göbel mirrors in the incident tion patterns of the homogenized foils are presented in
beam. The carburized foils were mounted on carbon Figure 1b. Evidently the homogenization treatment
tape and placed in the instrument. The positions of the results in symmetric peaks and the disappearance of
Bragg peaks in the X-ray diffractograms were obtained the peak associated with the carbon-free substrate, indi-
by fitting a pseudo-Voigt peak shape function through cating that redistribution and levelling of the carbon
the measured intensities forming the peak. In the fitting atoms over the thickness of the foil has indeed occurred.
procedure corrections for Ka2, Lorentz polarization and In Figure 2, the lattice parameter calculated from the
absorption factors were included. Due to a strong crys- spacing of the cC (2 2 0) lattice planes is presented as a
tallographic texture in the foils, the lattice parameters function of the number of carbon atoms per metal atom.
were determined from the position of the cC (2 2 0) This is equivalent to the occupancy of the octahedral
reflections. interstices, presuming that C is exclusively present at
X-ray diffraction patterns of the as-carburized AISI these locations. Evidently, a linear relation applies for
316 foils and the austenitized reference are given in Fig- the dependence of the lattice parameter, a, on the car-
ure 1a. Due to a strong crystallographic texture in the bon content, yC. The uncertainty in the peak position
foils, only the cC 2 2 0 line profile was available for inter- of the 2 2 0 reflection as obtained from fitting the pseu-
pretation. A shift of face-centered cubic (fcc) peaks to do-Voigt line shape is smaller than the symbols used in
lower diffraction angles is observed for all carburized Figure 2.
specimens. This is an immediate consequence of the Literature data for the lattice expansion caused by
expansion of the fcc lattice due to the dissolution of carbon in austenite as derived from the lattice expansion

a b 2500 reference
300
8
powder
yC=0.15
250
2000
3
relative intensity, I/Imax *10

yC=0.14
intensity reference, cps

6
log (Intensity [a.u.])

yC=0.11 yC=0.15 200

1500
yC=0.14
4 yC=0.09 150
1000 yC=0.11
yC=0.07
yC=0.09 100
2 yC=0.04
500 yC=0.07
50
reference powder yC=0.04
0 0
115 120 125 130 135 115 120 125 130 135
scattering angle, 2θ scattering angle, 2θ

Figure 1. X-ray diffraction patterns of the carburized (a) and homogenized (b) AISI 316 foils with six different carbon contents in the range
yC = 0.04–0.15. An austenitized reference powder is also included. Note that a logarithmic scale applies for (a), while a linear scale applies for (b).
 Author's personal copy
T. S. Hummelshøj et al. / Scripta Materialia 63 (2010) 761–763
Table 1. Lattice expansion by C dissolution and molar volume of C in austenite for stressed and unstressed retained austenite (Fe–C) and expanded
austenite (Fe–Cr–Ni–C). The lattice expansion coefficient is given in nm/wt.%, except where indicated with yN or yC, where the slope of the lattice
parameter vs. the occupancy of the octahedral interstices is given.
System Lattice expansion coefficient
Fe–Cr–Ni–N 0.0597 (yN)
Fe–Cr–Ni–C 0.0603 (yC)
Fe–Cr–Ni–C 0.0028
Fe–Cr–Ni–C 0.0028
Fe–C 0.0033
Fe–Cr–Ni–C 0.0041
Fe–C 0.0046
Fe–Cr–Ni–C 0.0054
of retained austenite with increasing carbon content ap-
ply for carbon contents up to yC = 0.12 [5] and are col-
lected in Table 1; the lattice expansion due to nitrogen in
expanded austenite [4] is also included for comparison.
In general, the literature data relating the carbon con-
tent in austenite to the lattice parameter is derived from
retained austenite in as-quenched martensitic specimens.
Ridley et al. [10] suggested that the smaller lattice
parameters observed for retained austenite, particularly
in the lower carbon content range, where the amount of
retained austenite is small, could be due to (hydrostatic)
compressive strains arising from the martensitic trans-
formation. Cheng et al. [11] suggested that stresses in
the majority of the martensite induce pronounced com-
pressive stresses in the retained austenite in Fe–C speci-
mens, corresponding to a compressive strain of about
0.4% in retained austenite. The data in Table 1 can be
approximately subdivided in two classes: (i) large lattice
expansion coefficient of carbon in strained retained aus-
tenite and (ii) low lattice expansion coefficient of carbon
in unstrained austenite. The reported values for lattice
expansion of austenite due to carbon dissolution range
from 0.0028 (nm/wt.% C) as determined in the present
work and by Kahn et al. [5] for unstrained austenite,
to the highest value determined from carburized speci-
mens without accounting for stress effects: 0.0054 (nm/
wt.% C) [2]. In order to obtain the dependence of the lat-
tice parameter of austenite on carbon content for un-
strained austenite, Cheng et al. [11] included data for
Fe–C specimens measured at elevated temperature,
where the specimens are austenitic (and thus strain-free),
and extrapolated these values to room temperature by
correcting for thermal shrink. Both dependencies,
strained and unstrained, are included in Table 1. Simi-
larly, in the most recent study by Kahn et al. [5] the
dependencies for strained expanded austenite and, after
correcting for stress effects, unstrained expanded austen-
ite, are included. Comparison of the data from the pres-
ent work with those published in the literature clearly
shows a much better agreement with relations for un-
strained austenite than for strained austenite. The pres-
ent work assesses for the first time the dependence of the
lattice parameter of expanded austenite simply on a
series of stress-free samples with different carbon
contents at room temperature without corrections for
stresses or thermal shrink.
A comparison of the molar volume for C and N in ex-
panded austenite [4], i.e., 4.20 and 4.15  106 m3 mol1,
respectively, shows a striking agreement. Nevertheless,
View publication stats
763
Composition range of validity yJ (J = N,C) Reference Stressed
0–0.61 [4] No
0–0.156 Present work No
0–0.156 Present work No
0–0.12 [5] No
0–0.09 [11] No
0–0.12 [5] Yes
0–0.09 [11] Yes
0–0.10 [2] Yes
the composition ranges where carbon and nitrogen stabi-
lize expanded austenite are different. Carbon is dissolved
for carbon contents up to yC = 0.156, while nitrogen is
dissolved for yN > 0.17 [12]. The reason for this difference
is the stronger binding of N to Cr than C to Cr and the
occurrence of short-range ordering of Cr and N in
Fe–Cr–Ni–N [12].
The lattice parameter, a, of expanded austenite was
for the first time determined experimentally in un-
strained austenite at room temperature from a range
of stress-free carburized foils of uniform composition,
and assessed to be linearly dependent on the occupancy
of the octahedral interstices, yC, according to:
a ¼ 3:5965½Å þ ð0:6029  0:0189Þy C or as
a ¼ 0:35965 ½nm þ ð0:0028  9:07  105 ÞxC ðwt:% CÞ;
which is in good agreement with existing literature data
for unstrained ferrous austenite as well as with nitrogen
in expanded austenite.
Financial support from the Danish Research
Council for Technology and Production Sciences under
Grant 274-05-0367 is gratefully acknowledged.
[1] Z.I. Zhang, T. Bell, Surf. Eng. 1 (1985) 131.
[2] Y. Sun, X. Li, T. Bell, Mater. Sci. Technol. 15 (1999)
1171.
[3] T. Christiansen, M.A.J. Somers, Scripta Mater. 50 (2004)
35.
[4] T.L. Christiansen, M.A.J. Somers, Metall. Mater. Trans.
A 37 (2006) 675.
[5] H. Kahn, G. Michal, F. Ernst, A.H. Heuer, Metall.
Mater. Trans. A 40 (2009) 1799.
[6] T.L. Christiansen, T.S. Hummelshøj, M.A.J. Somers,
Surf. Eng. 26 (2010) 242 (also published as Surface
Modification Technologies XXII, in: Proceedings of the
22nd International Conference on Surface Modification
Technologies, 2009, pp. 199–206).
[7] M.A.J. Somers, T. Christiansen, Patent Application WO
2006 136166, 2006.
[8] T.S. Hummelshøj, T.L. Christiansen, M.A.J. Somers,
Defect Diffus. Forum 273–276 (2008) 306.
[9] T.L. Christiansen, M.A.J. Somers, Int. J. Mater. Res. 99
(2008) 999.
[10] N. Ridley, H. Stuart, L. Zwell, Trans. Metall. Soc. AIME
245 (1969) 1834.
[11] L. Cheng, A. Böttger, Th.H. de Keijser, E.J. Mittemeijer,
Scripta Metall. Mater. 24 (1990) 509.
[12] T.L. Christiansen, M.A.J. Somers, Int. J. Mater. Res. 100
(2009) 1361.