Corrosion Science 51 (2009) 876881

Contents lists available at ScienceDirect

Corrosion Science
journal homepage: www.elsevier.com/locate/corsci

Microstructure, heat treatment and pitting corrosion of 13CrNiMo plate

and weld metals
P.D. Bilmes a, C.L. Llorente a, C.M. Mndez b, C.A. Gervasi c,d,*
a

Laboratorio de Investigaciones de Metalurgia Fsica (LIMF), Facultad de Ingeniera, Universidad Nacional de la Plata, calle 1 y 47, 1900 La Plata, Argentina
Laboratorio de Corrosin, Facultad de Ciencias Exactas, Qumicas y Naturales, Universidad Nacional de Misiones, Felix de Azara 1552, 3300 Posadas, Argentina
c
Instituto de Investigaciones, Fisicoqumicas Tericas y Aplicadas (INIFTA), Facultad de Ciencias Exactas, Universidad Nacional de la Plata, Sucursal 4-C.C.16, 1900 La Plata, Argentina
d
Laboratorio de Ingeniera de Corrosin y Tecnologa Electroqumica (LICTE), Facultad de Ingeniera, Universidad Nacional de la Plata, 1 y 47, 1900 La Plata, Argentina
b

a r t i c l e

i n f o

Article history:
Received 12 November 2008
Accepted 27 January 2009
Available online 4 February 2009
Keywords:
A. Stainless steel
B. Polarization
B. X-ray diffraction
C. Pitting corrosion
C. Welding

a b s t r a c t
Potentiodynamic measurements together with SEM were used to analyze susceptibility to pitting corrosion of 13CrNiMo plate metals. Volume fractions of austenite in tempered conditions affect pitting resistance. Characteristic potentials (critical pitting and repassivation) increase according to the retained
austenite content. Results correspond to a structural renement resulting from a double tempering procedure. A comparison is made with results from weld metal samples. The importance of this evaluation is
to be able to determine pitting resistance, either in a service application, or in laboratory test programs.
2009 Elsevier Ltd. All rights reserved.

1. Introduction
In recent years steel manufacturers have made available a range
of low carbon martensitic stainless steels that exhibit improved
weldability and corrosion resistance in sour and CO2-containing
media and offer an economical choice of materials between carbon
steels and duplex stainless steels. As a result, since then these
materials have experienced an ever growing application as welded
tubular products, particularly for the oil and gas industry [13].
However, precautions must be taken when high resistances to pitting corrosion in aqueous chloride environments and to sulphide
stress corrosion cracking are required.
In most fusion welding processes, solidication or phase transformations during thermal cycles result in chemical inhomogeneity and changes in metallurgical structures in both the weld
metal and parent metal adjacent to the weld. In turn these properties are directly related to the susceptibility to localized corrosion.
Moreover, since ller composition normally matches the parent
metal it is advisable to study comparatively the inuence of thermal cycles on localized corrosion for 13CrNiMo stainless steel weld
and parent metals.
* Corresponding author. Address: Instituto de Investigaciones, Fisicoqumicas
Tericas y Aplicadas (INIFTA), Facultad de Ciencias Exactas, Universidad Nacional de
la Plata, Sucursal 4-C.C.16, 1900 La Plata, Argentina. Tel.: +54 221 425 7430; fax:
+54 221 425 4642.
E-mail addresses: [email protected], [email protected] (C.A.
Gervasi).
0010-938X/$ - see front matter 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.corsci.2009.01.018

Accurate measurement of localized corrosion susceptibility is

essential to predict the life in service of the most critical components of a metallic structure, namely, their welded parts. Localized
corrosion measurements also become very important in evaluating
the suitability of welding procedures. In addition and considering
that pitting is an electrochemical process it becomes clear that
the best methods to study susceptibilities to this form of corrosion
are the electrochemical techniques.
Recently Bilmes et al. [4] reported potentiodynamic anodic
polarization measurements carried out on weld metal 13CrNiMo
stainless steel in chloride containing solutions in order to assess
the inuence of applied post weld heat treatments (PWHTs) on pitting corrosion resistance. They showed that characteristic potentials (pitting potential EP and repassivation potential ER)
increased according to the retained austenite content. In this work
we aim to gain a further insight into this question as well as into
the thermal effects of welding on parent metal by comparing results obtained with 13CrNiMo plate and weld metals. Safe welding
procedures should result in welds with corrosion properties meeting comparable requirements equal to the parent metal.
In this work we followed the same methodology as in our previous paper [4]. This methodology included the same test conditions as well as identical characterization of test specimens. Only
the test material was altered from weld metal (previous study)
to plate metal (current study). This was necessary not simply because the methodology had already proved successful but mainly
because this work aims at performing a comparative assessment

877

P.D. Bilmes et al. / Corrosion Science 51 (2009) 876881

of the results. More important is the fact that this comparative

assessment is highly relevant [57]. Thus, the study of welding
metallurgy requires a consideration of the following:
(a) a welded joint should normally resist corrosion the same as
the metal being joined;
(b) variations in composition or metallurgical or mechanical
changes brought about by welding can introduce corrosion
problems because corrosion thrives on microstructural differences adjacent to the base metal;
(c) when subjected to a corrosive environment the weld metal
may corrode more, or less as compared to the base metal.
Consequently, using appropriate heat treatments to simulate
the thermal cycles in this work we performed a comparative
evaluation of localized corrosion susceptibility of low carbon
13CrNiMo stainless steel weld and parent metals.
2. Experimental
2.1. Materials, specimen preparation and characterization
Specimens were cut from 13CrNiMo plate metals and they were
subjected to different heat treatments (HTs). The composition of
the plate metal is shown in Table 1, and a summary of the heat
treatments and nomenclature used along the text are shown in
Table 2.
For each condition, volume fractions of the austenite were measured by X-ray diffraction from a Rietveld analysis [8]. X-ray diffraction patterns were obtained at room temperature with a
Philips PW 1710 diffractometer, furnished with diffracted beam
graphite monochromator. Data were collected by using Cu-Ka
radiation in the range 10 6 2h 6 120 at a step interval of 0.02.
Rietveld analysis was performed by using the Fullprof program
[9]. According to this analysis, double tempered conditions resulted in higher amounts of retained austenite content, Table 2.
2.2. Pitting examination
Cyclic potentiodynamic measurements (sweeping rate 0.5
mVs1) were performed in a conventional three-electrode cell.

Table 1
Nominal composition (wt.%).
Grade

Mn

Si

Cr

Ni

Mo

13CrNiMo

0.03

0.6

0.3

12.74

3.71

0.53

0.01

0.008

Specimens cut from plate metal and subjected to different HTs as

indicated in Table 2 were used as working electrodes. These electrodes exhibiting 1 cm2 apparent area were ground to a mirror-like
nish before the application of the potential perturbation. A mirror-like nish was obtained after initial grinding with grit paper
and diamond paste and subsequent electropolishing in a solution
of HClO4 (62 ml), methanol (700 ml), butyl cellosolve (100 ml)
and H2O (137 ml). The counter-electrode was a large Pt sheet area.
All electrode potentials were measured and referred to in the text,
tables and gures against a Hg/Hg2SO4, K2SO4 (satd) reference
electrode (0.65 V in the NHE scale).
Experiments were carried out at room temperature (25 2 C)
in 0.05 M K2SO4 + 0.04 M NaCl (pH = 4.1) solution made from analytical grade reagents and triple-distilled water. Test solution composition was originally proposed in the literature [10] for this kind
of stainless steels and specied by the real environment.
A nitrogen stream was purged through the electrolyte solutions
to eliminate dissolved oxygen before each measurement. Inert gas
circulation was maintained above the solution level during the
measurements.
Optic and scanning electron microscopy was applied to observe
relatively large, already stable pits (diameter > 10 lm).
3. Results and discussion
The microstructures of all specimens consist of a typical lath
martensite with different content of retained austenite particles
precipitating either at grain boundaries or inside the martensite
grains (both in lath borders and packet borders). The morphology
of retained austenite corresponds to particle-platelet precipitates
with thickness smaller than 1 lm (Fig. 1).
Carrying out subcritical tempering below 600 C (E condition)
renders a ferritic structure that does not contain a relevant amount
of retained austenite (Fig. 2).
The intercritical tempering at 600 C (B and G conditions)
produces martensite decomposition together with precipitation
of a very thin austenite dispersion, since it is known that the precipitation of austenite takes place at tempering temperatures
slightly higher than Ac1.
In the microstructures with double tempering a structural
renement can be produced. This corresponds to higher austenite
content together with a more uniform distribution of this phase
(H, M and P conditions). Fig. 3 shows this result for M
condition.
X-ray diffraction patterns for M and E conditions are shown
in Figs. 4 and 5, respectively and can be comparatively analyzed.
Fig. 4 corresponds to a sample with retained austenite resulting
from double tempering at 670 + 600 C, while Fig. 5 corresponds

Table 2
Description of the applied heat treatments.
Condition

Heat treatments

Temperature (C)

Time (h)

Austenite content (vol %)

As-received
B
E

Annealing
Single tempering
Solution annealing + tempering

600
950 + 550

6
16
2

Solution annealing + tempering

950 + 600

Solution annealing + double tempering

950 + 670 + 600

Solution annealing + double tempering

950 + 670 + 600

Double tempering

670 + 600

2
1
2
1
2
1
2
2
1
2
8
2
2

8
18

22

20

878

P.D. Bilmes et al. / Corrosion Science 51 (2009) 876881

Fig. 1. Austenite particles (white particle-platelets) in a ferrite matrix (gray/black

areas) for 13CrNiMo plate metals.

to a sample without retained austenite resulting from single tempering at 500 C.

Pitting occurs on thin lms on stainless steel formed by sputtering deposition, i.e. free from non-metallic inclusions etc. [11,12].
Accordingly, it can be stated that breakdown of the passive lm
is one of the origins of pit initiation [1315]. In turn, oxide layer
composition varies according to the underlying microstructure.
For stainless steels, heat treatment, grinding, and abrasive blasting are detrimental to pitting resistance, whereas passive lms
formed after pickling in HNO3 + HF or passivation in HNO3 are benecial [16]. Heat treatments in air generate a chromium oxide scale
and a chromium-depleted region under the scale. The scale is typically removed mechanically, and the chromium-depleted region is
removed by pickling [16]. Seminal work in the literature related to
pitting corrosion deals with such aspects as the origins of pitting
corrosion on stainless steels as well as the effects of electrolyte
composition [17,18].
Electrochemical studies of pitting corrosion have shown that
characteristic potentials exist. Stable pits form at potentials noble
to the pitting potential EP, and will grow at potentials noble to
the repassivation potential, ER, which is lower than EP [19].
A typical cyclic potentiodynamic anodic polarization curve is
shown in Fig. 6, for a sample according to preparation condition B.
Since EP values exhibit the typical scatter for a pitting process, at
least 20 identical sweeps were performed for each condition, i.e.
for each different heat treatment, as indicated above. The large
set of EP values, obtained for each condition, was analyzed according to the usual stochastic approach considering the probability for
pitting P(E) [20,21]

PE n=1 N

Fig. 2. Ferrite matrix for 13CrNiMo plate metals with applied heat treatment
according to the E condition.

Fig. 3. Austenite particles in a ferrite matrix for 13CrNiMo plate metals with
applied heat treatment according to the M condition.

where N is the total number of samples studied, and n is the number

of samples that were pitted at a potential of E or lower. The potential at P = 0.5 was considered as the EP value, representative for the
material and its preparation condition. Fig. 7 shows the derived pitting probability as a function of E for a sample according to preparation condition B. The potential at P = 0.5 is indicated in this gure.
With regard to ER, these values vary linearly with the logarithm
of the charge density q passed under the hysteresis during the
sweep [22]. The potential for an arbitrarily selected amount of
pit propagation (q = 0.1 C cm2) was considered here as a normalized ER value to be used for comparison of the different sample
preparations.
Figs. 8 and 9 show pitting potentials EP and repassivation potentials ER, respectively, plotted against the amount of retained austenite for the different preparation conditions.
The linear dependence found for both EP and ER, conrms previous results obtained with weld metal samples [4]. The shift in the
noble direction according to the retained austenite content, determines an enhanced pitting resistance.
Enhanced pitting resistance corresponds to double tempering
PWHT conditions. These results indicate that, under present test
conditions, pitting resistance can be strongly affected by two
microstructural factors, namely, the amount and size of carbides
and the amount of retained austenite [4].
These results can be understood in terms of a structural renement resulting from two-stage tempering according to the model
previously advanced [23,24]. Single tempering at 600 C renders
a microstructure composed of tempered martensite and some retained austenite. Double tempering, initially at 670 C and then
at 600 C, generates the precipitation of new stable austenite particles-platelets through newly formed interfaces (fresh martensite/
austenite), while fresh martensite is decomposed into tempered
martensite. Most notably, longer tempering times at 600 C (8 h
for condition M) promotes a higher content of retained austenite
than for the other double-stage tempered conditions (2 h for

P.D. Bilmes et al. / Corrosion Science 51 (2009) 876881

879

Fig. 4. X-ray diffraction patterns of a sample with applied heat treatment according to the M condition.

Fig. 5. X-ray diffraction patterns of a sample with applied heat treatment according to the E condition.

conditions H and P). Hence, as a result of this double tempering,

the number of austenite platelets increase and they present a more
uniform distribution. Moreover, because the transformation is
diffusional the austenite is enriched in elements such as nickel,

carbon and nitrogen, what in turn determines dissolution or renement of carbide or carbonitride particles. The whole process generates a lower susceptibility to pitting corrosion [4 and references
therein]. As in the case of surface-melted martensitic stainless steel

880

P.D. Bilmes et al. / Corrosion Science 51 (2009) 876881

Fig. 6. Cyclic potentiodynamic polarization curve recorded in de-aerated 0.05 M

K2SO4 + 0.04 M NaCl (pH = 4.1), sweep rate 0.5 mVs1, for preparation condition B.
Pitting and repassivation potentials are indicated in the gure.

Fig. 7. Plot of the cumulative probability P for pitting potential values (Eq. (1)) as a
function of potential, for preparation condition B. The potential at P = 0.5 is
representative for the material and its preparation condition.

UNS S42000 the improvement in pitting corrosion resistance resulted from the dissolution or renement of carbide particles and
the presence of retained austenite, as evidenced by the fact that
pitting and repassivation potentials increased linearly with the
amount of retained austenite [4,25].
A comparison between the results in Figs. 8 and 9 and those
previously obtained for weld metal samples [4], shows that both
EP and ER values are always more noble for parent (or base) metal samples than for weld metal samples, for each volume fraction of retained austenite. However, the difference in EP values
DEP between the parent metal and the weld metal samples decreased with increasing the retained austenite content, while
for contents of P20 wt.% DEP becomes negligible. This nding
supports the viewpoint that the benecial condition of the parent metal (having a more homogeneous structure, with lower
density of inclusions and other defects acting as pit initiation
sites) becomes less prevailing compared to the benecial effect
of a high retained austenite content (associated with structural
renement and an increase in dissolved Cr content [26]). The difference in ER values (DER) decreases with increasing retained

Fig. 8. Retained austenite (wt.%) vs. pitting potential (EP) for different preparation
conditions as indicated in the text.

Fig. 9. Retained austenite (wt.%) vs. repassivation potential (ER) for different
preparation conditions as indicated in the text.

austenite content in a much less marked way. Thus, the relative

inuence of austenite content (measured as the ratio of slopes
dE/d[%aust]) is 2.13 for EP while only 1.25 for ER. This reects
the fact that the benecial effect of retained austenite on the
repassivation kinetics imposing a stop of growth on stably growing pits (related to ER) is not as strong as on the presence of pit
initiation sites (related to EP).
Pit size and shape depend on the selected experimental conditions. In the electrochemical tests regular-shaped pits are generated with hemispherical morphology (Fig. 10). Consequently,
these pits grow without following the crystallographic orientation
of the metal and cross the interface of crystallites. It has been stated that this is the result of the presence of sulfate in the electrolyte [13].
4. Conclusion
The presence and amount of retained austenite as a microstructural component resulting from the applied heat treatments has a
benecial effect on the pitting corrosion resistance of both weld
and parent metal 13CrNiMo stainless steels. When retained

P.D. Bilmes et al. / Corrosion Science 51 (2009) 876881

Fig. 10. SEM micrograph showing hemispherical pits grown on 13CrNiMo plate
metal in 0.05 M K2SO4 + 0.04 M NaCl (pH = 4.1) solution.

austenite is controlled at 2025% both materials exhibit comparable pitting corrosion resistances, given not only by similar critical
potentials for stable pit growth, but also exhibiting analogous
repassivation kinetics.
Acknowledgements
Drs. Gervasi and Llorente are grateful to the Comisin de Investigaciones Cientcas y Tcnicas Buenos Aires for their positions as
members of the Career of Scientic Researcher.
References
[1] T.G. Gooch, Heat treatment of welded 13%Cr-4%Ni martensitic stainless steels
for sour service, Welding J. 74 (1995) 213223.
[2] A.W. Marshall, J.C.M. Farrar, Welding of ferritic and martensitic 13%Cr steels,
in: IIW Doc 1998, No. IX-H422-98, International Institute of Welding,
Cambridge, England, 1998, pp. 8.
[3] A. Turnbull, A. Grifths, Corrosion and cracking of weldable 13 wt-%Cr
martensitic stainless steels for application in the oil and gas industry,
Corrosion Eng. Sci. Technol. 38 (2003) 2150.
[4] P.D. Bilmes, C.L. Llorente, L. Saire Huamn, L.M. Gassa, C.A. Gervasi,
Microstructure and pitting corrosion of 13CrNiMo weld metals, Corrosion
Sci. 48 (2006) 32613270.

881

[5] P. Sathiya, S. Aravindan, R. Soundararajan, A. Noorul Haq, Effect of shielding

gases on mechanical and metallurgical properties of duplex stainless-steel
welds, J. Mater. Sci. 44 (2009) 114121.
[6] R.S. Huang, L. Kang, X. Ma, Microstructure and phase composition of a lowpower YAG Laser-MAG welded stainless steel joint, J. Mater. Eng. Perform. 17
(2008) 928935.
[7] P. Ferro, A. Tiziani, F. Bonollo, Inuence of induction and furnace postweld heat
treatment on corrosion properties of SAF 2205 (UNS 31803), Welding J. 87
(2008) 298306.
[8] B.D. Cullity, Elements of X-ray Diffraction, Addison-Wesley, MA, 1978.
[9] C. Rodrguez, FULLPROF, a program for Rietveld renement and pattern
matching analysis, in: Proc. Satellite Meeting on Powder Diffraction of
Congress of IUCr, Tolouse, France, 1990, pp.127.
[10] S. Hashizume, K. Masamura, I. Matsushima, Pitting corrosion resistance of 13%
Cr steel in oil eld environments, Trans. ISIJ 23 (1983) B351.
[11] D.E. Williams, R.C. Newman, Q. Song, R.G. Kelly, Passivity breakdown and
pitting corrosion of binary alloys, Nature 350 (1991) 216219.
[12] M.P. Ryan, N.J. Laycock, R.C. Newman, H.S. Isaacs, The pitting behavior of ironchromium thin lm alloys in hydrochloric acid, J. Electrochem. Soc. 145 (1998)
15661571.
[13] H. Strehblow, Pitting corrosion, in: A.J. Bard, M. Stratmann (Eds.), Encyclopedia
of Electrochemistry, in: M. Stratmann, G.S. Frankel (Eds.), Localised Corrosion
Phenomena, vol. 4, Wiley-VCH, Weinheim, Germany, 2003, pp. 308343.
[14] K. Sugimoto, in: Advances in Materials Research, in: Y. Waseda, S. Suzuki
(Eds.), Characterization of Corrosion Products on Steel Surfaces, vol. 7,
Springer, Berlin, Heidelberg, 2006 (Chapter 1).
[15] P. Schmuki, H. Bhni, Illumination effects on the stability of the passive lm on
iron, Electrochim. Acta 40 (1995) 775783.
[16] A.J. Sedriks, Corrosion of Stainless Steels, second ed., John Wiley, New York,
1996.
[17] P.C. Pistorius, G.T. Burstein, Aspects of the effects of electrolyte composition on
the occurrence of metastable pitting on stainless steel, Corrosion Sci. 36 (1994)
525538.
[18] G.T. Burstein, C. Liu, R.M. Souto, S.P. Vines, Origins of pitting corrosion,
Corrosion Eng. Sci. Technol. 39 (2004) 2530.
[19] G.S. Frankel, Pitting corrosion, in: S.D. Cramer, B.S. Covino Jr. (Eds.), Corrosion:
Fundamentals, Testing, and Protection, ASM Handbook, vol. 13A, ASM
International, 2003, pp. 236241.
[20] G.S. Frankel, Pitting corrosion of metals: A review of the critical factors, J.
Electrochem. Soc. 145 (1998) 21862198.
[21] T. Shibata, T. Takeyama, Stochastic theory of pitting corrosion, Corrosion 33
(1977) 243251.
[22] B.E. Wilde, Critical appraisal of some popular laboratory electrochemical tests
for predicting the localized corrosion resistance of stainless alloys in sea water,
Corrosion 28 (1972) 283291.
[23] C.A. Gervasi, P.D. Bilmes, C.L. Llorente, Metallurgical factors affecting localized
corrosion of low-C 13CrNiMo martensitic stainless steels, in: I.S. Wang (Ed.),
Corrosion Res. Trends, Nova Science Publishers, New York, 2007, pp. 134.
Chapter 1.
[24] P.D. Bilmes, M. Solari, C.L. Llorente, Characteristics and effects of austenite
resulting from tempering of 13Cr-NiMo martensitic steel weld metals, Mater.
Characterization 46 (2001) 285296.
[25] C.T. Kwok, H.C. Man, F.T. Cheng, Cavitation erosion and pitting corrosion
behaviour of laser surface-melted martensitic stainless steel UNS S42000, Surf.
Coat. Technol. 126 (2000) 238255.
[26] M. Kimura, Y. Miyata, T. Toyooka, Y. Kitahaba, Effect of retained austenite on
corrosion performance for modied 13% Cr steel pipe, Corrosion 57 (2001)
433439.