Effects of W substitution on the precipitation of secondary phases and the associated pitting corrosion in hyper duplex stainless steels

Effects of W substitution on the precipitation of secondary phases and the associated pitting corrosion in hyper duplex stainless steels

Journal of Alloys and Compounds 544 (2012) 166–172 Contents lists available at SciVerse ScienceDirect Journal of Alloys and Compounds journal homepa...

1MB Sizes 0 Downloads 6 Views

Journal of Alloys and Compounds 544 (2012) 166–172

Contents lists available at SciVerse ScienceDirect

Journal of Alloys and Compounds journal homepage: www.elsevier.com/locate/jalcom

Effects of W substitution on the precipitation of secondary phases and the associated pitting corrosion in hyper duplex stainless steels Soon-Hyeok Jeon a, Soon-Tae Kim a, In-Sung Lee a, Ji-Soo Kim b, Kwang-Tae Kim b, Yong-Soo Park a,⇑ a b

Department of Material Science and Engineering, Yonsei University, 134 Shinchon-dong, Seodaemun-gu, Seoul 120-749, Republic of Korea Stainless Steel Research Group, POSCO Technical Research Laboratories, Goedong-dong, Nam-Gu, Pohang, Gyeongbuk 790-785, Republic of Korea

a r t i c l e

i n f o

Article history: Received 27 June 2012 Received in revised form 24 July 2012 Accepted 26 July 2012 Available online 4 August 2012 Keywords: Metals and alloys Corrosion X-ray diffraction Scanning electron microscopy

a b s t r a c t To elucidate the effects of W substitution on the precipitation of secondary phases and associated pitting corrosion in hyper duplex stainless steels, potentiodynamic polarization test, a SEM–EDS analysis and thermodynamic calculation were conducted. W substitution for Mo to the base alloy reduces the total amount of secondary phases. Particularly, W substitution alloy results in pronouncedly suppressing the amount of sigma phase whereas it slightly increases chi phases. W substitution alloy retards the deterioration of resistance to pitting corrosion by the precipitation of secondary phases due to delayed precipitation of the secondary phase. Experimental results are compared with calculations of alloying elemental activities. The thermodynamic calculations for the precipitation of secondary phases were in good agreement with the experimental results. Ó 2012 Elsevier B.V. All rights reserved.

1. Introduction Duplex stainless steels have been increasingly used for various applications such as power plants, desalination facilities and chemical plants due to high resistance to crevice and pitting corrosion with a high strength, good resistance to stress corrosion cracking (SCC), a relatively low cost, compared with other higher performance materials such as super austenite stainless steels [1–3]. However, in heat exchanger application, corrosion resistance of super duplex stainless is insufficient for higher temperature service or for a long service life, and where materials, with even higher corrosion resistance, are needed. Hyper duplex stainless steels (HDSS) with a PREW (Pitting Resistance Equivalent with Tungsten = wt.% Cr + 3.3(wt.% Mo+ 0.5 wt.% W) + 30 wt.% N) value above 50 were developed to meet industry demands for higher operating temperatures and longer run times. The existing researches were progressed for DSS and super DSS of PREW 40 grades. However, few studies have been focused on HDSS of PREW 50 grades, having higher Cr, Mo and W than DSS and super DSS, to substitute the 6% Mo austenite stainless steels in more severe environment. The secondary phases such as sigma (r) phase and chi (v) phase are the intermetallic phases that forms at high temperature 600–950 °C [4,5]. These intermetallic phases lower the fracture toughness and the corrosion resistance significantly [6,7]. The intermetallic compounds affect not only the mechanical and corrosion properties, but also the thermoelectric and magnetic ⇑ Corresponding author. E-mail address: [email protected] (Y.-S. Park). 0925-8388/$ - see front matter Ó 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jallcom.2012.07.129

properties of duplex stainless steels. Lara et al. [8] reported that the precipitation of sigma phases reduce the thermoelectric property due to the reduction of ferrite content at 650–900 °C in duplex stainless steels. Lo et al. [9] showed that magnetic susceptibility decreases due to the decrease of ferromagnetic phase during spinodal decomposition of ferrite phases in duplex stainless steels at 650–800 °C. For duplex stainless steels, the tendency to secondary phases precipitation is crucial since the existence of the ferrite phase will enhance the kinetics for precipitation of secondary phases. The sigma phase preferentially precipitates into the ferrite due to the higher Cr and Mo concentration in the ferrite phase [10]. A fundamental reason why the sigma phase preferentially grows into the ferrite phase is that the ferrite phase is thermodynamically metastable at temperature where the sigma phase precipitates [11]. The retardation of the formation the r phase and v phase through the partial substitution of W for Mo has been investigated. Ogawa et al. [12] showed that the addition of tungsten (W) in the heat affected zone (HAZ) of duplex stainless steel is effective in delaying sigma phase precipitation due to the enhancement the tendency to form the chi phase with a higher nucleation efficiency and lower growth rate than the sigma phase. Kim and Kwon [13] showed that a partial substitution of W for Mo in 25% Cr duplex stainless steels retarded the precipitation of sigma (r) phase in the alloys. Park et al. [14] reported that a partial substitution of W for Mo in 29% Cr ferritic stainless steels not only increased resistance to pitting corrosion of the alloys by a synergic contribution of Mo and W but also retarded the amount of precipitation of sigma (r) phase in the alloys.

167

S.-H. Jeon et al. / Journal of Alloys and Compounds 544 (2012) 166–172 Table 1 Chemical compositions of the experimental alloys (wt.%). Alloy

C

Cr

Ni

Mo

W

Si

Mn

S

N

Fe

4.6 Mo 2.8 Mo–3.6 W

0.023 0.023

26.83 27.21

6.71 7.05

4.64 2.76

– 3.58

0.29 0.29

0.97 0.95

0.0040 0.0037

0.39 0.36

Bal. Bal.

a

b 0.8

0.8

0.7

ferrite 0.6

austenite

0.5

sigma

0.4

ferrite

0.3 0.2

chi 0.1

Mole fraction of phases

Mole fraction of phases

0.7

austenite

0.6

ferrite 0.5 0.4

sigma ferrite

0.3 0.2

chi

0.1

HCP

0

HCP

0 500

600

700

800

900

1000 1100 1200

500

ο

600

700

800

900

1000 1100 1200 ο

Temperature ( C)

Temperature ( C)

Fig. 1. Mole fraction of stable phases in the experimental alloys calculated using TCFE5 (Thermo-Calc calculation): (a) 4.6 Mo alloy and (b) 2.8 Mo–3.6 W alloy.

Fig. 2. Optical micrograph of the experimental alloys after solution heat treatment at 1090 °C: (a) 4.6 Mo alloy and (b) 2.8 Mo–3.6 W alloy.

4.6 Mo

2.8 Mo-3.6 W

Aged for 10 min o at 850 C

50 μm

Aged for 30 min o at 850 C

Fig. 3. Back-scattered electron (BSE) images of the micrograph of the experimental alloy aged at 850 °C for 10 min and 30 min.

168

S.-H. Jeon et al. / Journal of Alloys and Compounds 544 (2012) 166–172

Table 2 Chemical composition of the secondary phases formed in experimental alloys after aging at 850 °C for 10 min (wt.%). Phase

4.6 Mo

Ferrite (a) Austenite (c) Chi (v) Sigma (r)

2.8 Mo–3.6 W

Fe

Cr

Mo

Fe

Cr

Mo

W

56.0 5.72 51.7 54.7

29.1 25.8 27.6 30.2

6.2 3.8 15.8 9.3

55.8 58.0 45.4 54.2

28.7 25.9 25.5 29.1

3.2 2.0 10.9 4.6

4.2 2.6 13.3 5.5

In the previous study, two possible mechanisms for partial substitution of W for Mo were suggested. The precipitation of v phase seems to be closely associated with the retardation of the precipitation r phase. Since the formation of r phase in the alloys requires high concentrations of Mo and W, the preferential precipitation of v phase in the alloy during the initial period of aging can inhibit the nucleation and growth of r phase by depleting W and Mo adjacent to the v precipitates. In addition, the retardation of the growth of the sigma and chi phases in W added alloy may be attributed to the inherent difference in diffusion rate between W and Mo. It was reported that the diffusion rate of W at 850 °C was 10–100 times slower than that of Mo in ferrous alloys [15].

a

500

(111)

Intensity (cps)

400

5Mo 4.6 Mo 2.5Mo-3.5W 2.8 Mo-3.6 W

(110)

2.1. Material and heat treatment The alloys used in this work were melted in a high frequency vacuum induction furnace and then hot rolled to plates of 6 mm thickness. The chemical compositions of the alloys are presented in Table 1. The alloys were designed to have chemical compositions of Fe–27Cr–7Ni–4.6Mo–0.35N and Fe–27Cr–7Ni–2.8Mo–3.6W– 0.35N, respectively, with same PREW value (52). The alloys were named 4.6 Mo and 2.8 Mo–3.6 W according to their major composition. The experimental alloys were cut and solution heat-treated for 5 min per 1 mm thickness at 1090 °C and then quenched in water. The specimens were then isothermally aged at 850 °C for 10 min and 30 min.

The secondary phases were observed using scanning electron microscopy (SEM) in backscattered electron mode (BSE). In addition, the chemical compositions of the secondary phases were analyzed by an energy dispersive spectroscopy (EDS) attached to the SEM. X-ray diffraction was performed on specimens containing a relatively large fraction of precipitates for the phase identification.

300

(200) 200

(212) (411)

(112)

2.3. Potentiodynamic anodic polarization test

(331)

0 40

42

44

46

48

50

52

2θ (ο) 500

4.6 Mo 2.8 Mo-3.6 W

(111) 400

Intensity (cps)

2. Experimental procedures

2.2. Phase identification

100

b

In the present work, thermodynamic calculations are performed for duplex stainless steels. Li et al. [16] studied that models used to calculate materials properties including time–temperature transformation (TTT) diagrams and mechanical properties of duplex stainless steel. However, in terms of thermodynamics, the mechanisms of the effects of W substitution on the retardation of secondary phases and associated pitting corrosion in hyper duplex stainless steels have not been verified. Accordingly, it is necessary to predict the activity of alloying elements such as chromium (Cr), molybdenum (Mo) and tungsten (W) using Thermo-Calc software to verify the effects of W substitution on the mechanism of retardation of secondary phases and to investigate the behaviors of associated pitting corrosion for hyper duplex stainless steels. In this study, to elucidate the effects of partial substitution of W for Mo on the precipitation of secondary phases and the associated pitting corrosion in hyper duplex stainless steels, potentiodynamic anodic polarization test, a scanning electron microscope–energy dispersive spectroscope (SEM–EDS) analysis, thermodynamic calculation using Thermo-Calc software were conducted.

To analyze the effects of W substitution on the resistance to localized corrosion of the experimental alloys, a potentiodynamic anodic polarization test was conducted. To measure the pitting potential (Ep) and the passive current density of the alloys, the potentiodynamic anodic polarization test was conducted in a deaerated 4 M NaCl solution at 70 °C and a deaerated 0.5 M HCl + 1 M NaCl solution at 40 °C per the ASTM G 5 [17]. Test specimens were joined with copper wire through soldering (95 wt.% Sn–5 wt.% Sb), and then mounted with an epoxy resin. One side of the sample was ground to 600 grit using SiC abrasion paper. After defining the exposed area of the test specimen as 1 cm2, the remainder was painted with a transparent lacquer. The test was conducted at a potential range of 0.65 VSCE  + 1.1 VSCE and a scanning rate of 0.06 V/min, using a saturated calomel electrode. 2.4. Thermodynamic equilibrium calculation

300

To predict the effects of W substitution on the activity of alloying element such as chromium (Cr), molybdenum (Mo) and tungsten (W) in the alloys, thermodynamic calculation using Thermo-Calc software was conducted. The equilibrium fractions of each phase were calculated against the temperature for the alloy using a commercial Thermo-Calc software package. The steel database TCFE5 available with the Thermo-Calc software was used to perform the calculation and it should be stressed that such calculation give the equilibrium state of system.

(200) 200

(112) (110)

100

(212) (411) (331)

3. Results and discussion

0 40

42

44

46

48

50

52

2θ (ο) Fig. 4. X-ray diffraction patterns of the experimental alloy aged at 850 °C for (a) 10 min and (b) 30 min.

3.1. Calculation of the equilibrium fractions of each phase The equilibrium fractions of each phase against the temperature for the HDSS alloy were calculated using commercial Thermo-Calc

169

S.-H. Jeon et al. / Journal of Alloys and Compounds 544 (2012) 166–172

60 55 50 45 40 35 30 25 20 15 10 5 0

4.6 Mo 2.8 Mo-3.6 W

σ phase

b

30

4.6 Mo 2.8 Mo-3.6 W

25

Area Fraction (%)

Area Fraction (%)

a

χ phase

20 15 10 5 0

0

5

10

15

20

25

30

0

5

10

15

20

25

30

Aging time at 850 oC (min.)

Aging time at 850 oC (min.)

c

Fig. 5. Area fraction of secondary phases formed in the experimental alloy aged at 850 °C: (a) r phase, (b) v phase and (c) r + v phases.

Pitting potential (mV vs. SCE)

1200

4.6Mo 2.8Mo-3.6W

1000 800

phase increases with increasing tungsten content. Tungsten (W) is a strong stabilizer for the v phase. Park et al. [14] reported that the v phase has higher affinity for W than for Mo and the great affinity of v phase for W can affect the reduction of the formation of r phase in the ferritic stainless steels.

600

3.2. Microstructural analysis 400 200 0 -200 -400 0

5

10

15

20

25

30

o

Aging time at 850 C (min.) Fig. 6. Effect of aging on the pitting potentials of the experimental alloys determined from potentiodynamic polarization test in deaerated 4 M NaCl solution at 70 °C according to ASTM G 5.

software (Fig. 1) package. In the 4.6 Mo alloy, It can be seen that the equilibrium phases at 850 °C are ferrite (a), austenite (c), sigma (r), chi (v) and a HCP phase with a composition corresponding to Cr2N nitride (Fig. 1, (a)). However, in the 2.8 Mo–3.6 W alloy, it can be seen that the equilibrium phases at 850 °C are ferrite (a), austenite (c), sigma (r) and a HCP phase (Fig. 1, (b)). This difference in precipitation of chi phases in experimental alloys should be attributed to W substitution for Mo. According to the calculations the precipitation range of the chi phases extends up to approximately 805 °C for the 4.6 Mo alloy and 860 °C for the 2.8 Mo–3.6 W alloy. The higher the tungsten the higher the temperature limit for precipitation of chi phases. The amount of the chi

Fig. 2 shows the microstructures of the experimental alloys which were solution heat treated at 1090 °C for 30 min. The austenite phase can be found as island phase on the background of ferrite phase which looks relatively dark. No secondary phases were observed in either alloy. Besides, affected by hot rolling, the specimen has elongated texture parallel to the rolling direction. Fig. 3 shows back-scattered electron (BSE) images of the 4.6 Mo and 2.8 Mo–3.6 W alloys aged at 850 °C for 10 and 30 min. For the 4.6 Mo alloy aged for 10 min, a large amount of sigma phase was observed since almost all the ferrite phase was transformed. The sigma phases became widespread both along the ferrite (a)/austenite (c) phase boundaries and within grain of the ferrite phase. However, for the 2.8 Mo–3.6 W alloy aged for 10 min, the dominant secondary phase is the chi phase which precipitates mainly along the ferrite (a)/austenite (c) phase boundaries with a small amount of the sigma phase precipitating randomly within grain of the ferrite phase. After aging for 30 min at 850 °C, for the 4.6 Mo alloy, the amount of sigma phase is almost same to the that of the 4.6 Mo alloy aged for 10 min since already most of the ferrite phase was transformed during 10 min. For the 2.8 Mo–3.6 W alloy aged for 30 min, the amount of precipitated secondary phases at grain boundaries and within grains increased. Especially, sigma phases formed widely along ferrite (a)/austenite (c) phase grain boundaries as well as within grains of the ferrite phase. In presented in Table 2, the chemical composition of sigma and chi phase is conducted using SEM–EDS. This result suggests that tungsten

170

S.-H. Jeon et al. / Journal of Alloys and Compounds 544 (2012) 166–172

a

is much more than that of the 2.8 Mo–3.6 W alloy. This result suggests that W substitution effectively retards the precipitation of the secondary phases. The amount of the chi phase increases with increasing tungsten content. W should be a strong stabilizer for the v phase.

1200

4.6 Mo-S.T. 4.6 Mo-10min 4.6 Mo-30min

Potential (mV vs. SCE)

1000 800 600

3.3. Effect of secondary phases on the resistance to pitting corrosion

400

Sigma phase is an intermetallic compound enriched in Cr and Mo [18,19]. Due to its high Cr and Mo contents, the precipitation of sigma phase depletes the surrounding regions in Cr and Mo, which deteriorate the corrosion resistance [20,21]. In this study, the formations of the secondary phases result in Cr, Mo and W depleted zones adjacent to v phases and r phases. In order to investigate the effects of W substitution on the precipitation of secondary phases and associated localized corrosion resistance of hyper duplex stainless steels, the pitting potential of the alloys were determined form potentiodynamic anodic polarization curves measured in 4 M NaCl solution at 70 °C, and presented in Fig. 6. While the pitting potentials of the 4.6 Mo alloy decreased with aging, that of the 2.8 Mo–3.6 W alloy remained relatively high values in spite of aging. The decrease in the pitting potential of the alloys with aging is due to the precipitation of the sigma phase and the resultant depletion of Cr around sigma phase. On the other hand, the delay in the degradation of the resistance to pitting corrosion of the alloys with an increase of W content appears to be due to the retardation in precipitation rate of sigma phase by W. Fig. 7 shows the effect of secondary phases on the potentiodynamic anodic polarization behaviors of the 4.6 Mo and 2.8 Mo– 3.6 W alloys in a deaerated 0.5 M HCl + 1 M NaCl solution at 40 °C per ASTM G 5. In general, the pitting potential (Ep) is defined as the breakdown potential destroying a passive film. As the Ep of an alloy increases, the resistance to pitting corrosion of the alloy increases. After heat treatment at 1090 °C for 30 min, the resistance to pitting corrosion of both the 4.6 Mo and 2.8 Mo–3.6 W alloys is excellent due to an increase of current density above oxygen evolution potential. However, with increasing aging time, based upon a decrease in pitting potential, the resistance to pitting corrosion decreased. The resistance to pitting corrosion of the 2.8 Mo– 3.6 W alloy was superior to that of the 4.6 Mo alloy because the pitting potential of the 2.8 Mo–3.6 W alloy is much higher than that of the 4.6 Mo alloy due to the retardation of secondary phases precipitation.. The reason that the resistance to pitting corrosion with the tungsten (W) substitution for molybdenum (Mo) increased is as follow: Difference in the resistance to pitting corrosion between the 4.6 Mo alloy and the 2.8 Mo–3.6 W alloy can be attributed to the degree of the secondary phase precipitation. The formations of the secondary phases result in a depletion of chromium (Cr), molybdenum (Mo) and tungsten (W). Cr, Mo and W depleted regions lead to a decrease in pitting corrosion resistance. The amount of formed secondary phases of the 4.6 Mo alloy is more than that of the 2.8 Mo–3.6 W alloy. The deterioration of resistance to pitting corrosion by the precipitation of secondary phases was significantly retarded in the 2.8 Mo–3.6 W alloy compare with that in the 4.6 Mo alloy due to retardation of the sigma phase precipitation.

200 0 -200 -400 -600 -7

10

-6

-5

10

10

-4

10

-3

10

-2

10

Current Density(A /cm2)

b 1200

2.8 Mo-3.6 W-S.T. 2.8 Mo-3.6 W-10 min 2.8 Mo-3.6 W-30 min

Potential (mV vs. SCE)

1000 800 600 400 200 0 -200 -400 -600 -7

10

-6

10

-5

10

-4

10

-3

10

-2

10

Current Density(A /cm2) Fig. 7. Potentiodynamic polarization behavior of the experimental alloys in deaerated 0.5 M HCl + 1 M NaCl solution at 40 °C according to ASTM G 5: (a) 4.6 Mo alloy and (b) 2.8 Mo–3.6 W alloy.

(W) substitution retards the precipitation of the secondary phases effectively in the early stage of aging, but the effectiveness of W substitution decreases as aging time increases. The calculated precipitation ranges of sigma (r) phase and chi (v) phase correspond well with experimental results. Fig. 4 shows the X-ray diffraction patterns of the alloys aged at 850 °C for 10 and 30 min. With increasing aging time, the peak of ferrite phase is led to become lower conspicuously, gradually deriving sigma phase. After aging for 10 and 30 min, the peak of sigma phase of the 4.6 Mo alloy is higher than that of the 2.8 Mo–3.6 W alloy. Small amounts of sigma and chi phases are not detected by X-ray diffraction due to overlap between secondary phases with ferrite phase and, mainly, austenite reflections. Nevertheless, the sigma phase reflections such as (1 1 2), (2 1 2), (4 1 1) and (3 3 1) are led to become higher conspicuously. Fig. 5 shows the area fraction of sigma (r) phase, chi (v) phase and total secondary (r + v) phases in the experimental alloys aged at 850 °C by image analyzer. With an increase in aging time, the area fraction of precipitated sigma phase in the 4.6 Mo alloy is more than that of 2.8 Mo–3.6 W alloy (Fig. 5 (a)). However, the area fraction of formed chi phase in the 2.8 Mo–3.6 W alloy is more than that of 4.6 Mo alloy (Fig. 5 (b)). Fig. 5 (c) shows that the area fraction of total secondary phases (r + v) precipitated of 4.6 Mo alloy

3.4. Mechanism on the effect of W substitution on the precipitation of secondary phases Cao et al. [22] also suggested that the v phase had a lower nucleation barrier than the r phase, but only stable r phase survived after long aging. As noted in the previous metallographic observations, the v phase precipitates earlier than the r phase,

171

S.-H. Jeon et al. / Journal of Alloys and Compounds 544 (2012) 166–172

a

b

50 45

50 45

Cr

35

30

Activity

Activity

35

Fe

25 20

30

Fe

25 20

Mo

15

15

Ni

5

C

5 -4

-4

10

0 500

600

700

800

900

1000 1100 1200

W

Mo

10

10 10

Cr

40

40

Ni C

0 500

o

600

700

800

900

1000 1100 1200 o

Temperature ( C)

Temperature ( C)

Fig. 8. Activity of alloying element in the experimental alloys versus temperature using TCFE5 (Thermo-Calc calculation): (a) 4.6 Mo alloy and (b) 2.8 Mo–3.6 W alloy.

2.8 Mo- 3.6 W alloy

4.6 Mo alloy

Solution heat treatment at 1090 oC : 50 %

: 49 %

:4%

: 28 %

:1%

:6%

Aged at 850 oC for 10 min

Aged at 850 oC for 30 min

Pitting corrosion

• The total amount of the secondary phases : 4.6 Mo alloy > 2.8 Mo-3.6 W alloy • The resistance to pitting corrosion : 4.6 Mo alloy < 2.8 Mo-3.6 W alloy

Fig. 9. Schematic of the effects of W substitution on the precipitation of secondary phases and the associated pitting corrosion in the experimental alloys

and v phase appears to be meta-stable and act as a transition phase in the duplex stainless steels. Dobranszky et al. [23] reported that small amount of chi phase formed earlier than the r phase on the ferrite-ferrite boundaries, but later chi phase transformed into sigma phase. The preferential precipitation of v phase seems to be closely associated with the retardation of the precipitation r phase. Since the formation of r phase in the both alloys requires high concentrations of Mo and W, the preferential precipitation of v phase in the alloy during the initial period of aging can inhibit the nucle-

ation and growth of r phase by depleting W and Mo adjacent to the v precipitates. The addition of tungsten retards sigma phase precipitation. When tungsten is added, the chi phase as the quasi-equilibrium phase is preferentially nucleated than sigma phase because tungsten is a strong stabilizer for chi phase. The preferential precipitation of the chi phase during the early stage of aging reduces that the probability of the sigma phases being nucleated by depleting Mo and W along grain boundaries by the precipitation of chi phase. In addition, growth of the sigma phase is due to diffusion of tung-

172

S.-H. Jeon et al. / Journal of Alloys and Compounds 544 (2012) 166–172

sten in tungsten (W) added alloy. The diffusion rate of tungsten in the ferrous alloys is around 10–100 times slower than that of Mo [15]. This result shows that the growth of sigma containing tungsten delays in comparison with the tungsten free sigma phase. The effect of partial substitution of W for Mo on the activity of alloying element such as iron (Fe), nickel (Ni), chromium (Cr), molybdenum (Mo) and tungsten (W) can be calculated as a function of temperature as shown in Fig. 8. With the partial substitution of W for Mo to the base alloy, the activity of molybdenum (Mo) decreases and the activity of tungsten (W) increases. Molybdenum (Mo) and tungsten (W) is also the main element stabilizing thermodynamically the sigma and chi phases like Cr. Since tungsten (W) is a strong stabilizer for the v phase, the amount of the chi phase increases with increasing activity of tungsten. The preferential precipitation of the chi phase during the early stage of aging reduces that the probability of the sigma phases being nucleated by depleting Mo and W along grain boundaries by the precipitation of chi phase. Park et al. [14] reported that the v phase has higher affinity for W than for Mo and the great affinity of v phase for W can affect the reduction of the formation of r phase in the ferritic stainless steels. Mo is one of the key elements in deleterious secondary phases, the decrease of activity of Mo has an effect on the retardation of secondary phase precipitation. Since Mo is a strong stabilizer for the sigma phase, the W substitution for Mo to the base alloy results in pronouncedly suppressing the amount of sigma phase due to the decrease of activity of Mo. W substitution for Mo to the base alloy reduces the total amount of secondary phases due to the decrease of activity of molybdenum and increase of activity of tungsten. Particularly, W substitution for Mo to the base alloy results in pronouncedly suppressing the amount of sigma phase whereas it increases chi phases. In summary, the effect of W substitution on the precipitation of secondary phases and associated pitting corrosion in the experimental alloys is schematically presented in Fig. 9. With an increase in aging time, the total amount of secondary phases precipitated in the both alloys increases. Although the secondary phases precipitated became further grown in the both alloys, the precipitation rate and type of secondary phases is different from the 4.6 Mo alloy to the 2.8 Mo–3.6 W alloy. For the 4.6 Mo alloy aged for 10 min, a large amount of sigma phases formed widely within a-grains a large amount of sigma phase was observed since almost all the ferrite phase was transformed. However, for the 2.8 Mo–3.6 W alloy aged for 10 min, the dominant secondary phase is the chi phase which precipitates mainly along the ferrite (a)/austenite (c) phase boundaries with a small amount of the sigma phase precipitating randomly within grain of the ferrite phase. The W substitution to the base alloy results in pronouncedly suppressing the amount of sigma phase whereas it increases chi phases. The W substitution to the base alloy reduces the total amount of secondary phases. After aging for 30 min at 850 °C, for the 4.6 Mo alloy, the amount of sigma phase is almost same to the that of the 4.6 Mo alloy aged for 10 min since already all the ferrite phase was transformed during 10 min. For the 2.8 Mo–3.6 W alloy aged for 30 min, the amount of precipitated secondary phases at grain boundaries and within grains increased. Especially, sigma phases formed widely along ferrite (a)/austenite (c) phase grain boundaries as well as within grains of the ferrite phase. The deterioration of resistance to pitting corrosion by the precipitation of secondary phases was significantly retarded in the 2.8 Mo–3.6 W alloy compare with that in the 4.6 Mo alloy due to the delayed precipitation of the sigma phase. 4. Conclusions In this study, to elucidate the effects of W substitution on the precipitation of secondary phases and associated pitting corrosion

in hyper duplex stainless steels, potentiodynamic anodic polarization test, a scanning electron microscope–energy dispersive spectroscope (SEM–EDS) analysis, thermodynamic calculation using Thermo-Calc software were conducted. 1. The W substitution to the base alloy results in pronouncedly suppressing the amount of sigma phase whereas it increases chi phases. The W substitution to the base alloy reduces the total amount of secondary phases. This result suggests that the W substitution effectively retards the precipitation of the secondary phases. When tungsten is added, the chi phase as the quasi-equilibrium phase is preferentially nucleated than sigma phase because tungsten is a strong stabilizer for chi phase. 2. With the partial substitution of W for Mo to the base alloy, the activity of molybdenum (Mo) decreases and the activity of tungsten (W) increases. The amount of the chi phase increases with increasing activity of tungsten. The preferential precipitation of the chi phase during the early stage of aging reduces that the probability of the sigma phases being nucleated by depleting Mo and W along grain boundaries by the precipitation of chi phase. Since Mo is one of the key elements in deleterious secondary phases, the decrease of activity of Mo has an effect on the retardation of secondary phase precipitation. The W substitution for Mo to the base alloy reduces the total amount of secondary phases due to the decrease of activity of molybdenum and increase of activity of tungsten. 3. The W substitution for Mo to the base alloy retards the deterioration of resistance to pitting corrosion by the precipitation of secondary phases due to delayed precipitation of the secondary phase.

Acknowledgement This work has been supported by Ministry of Knowledge Economy of the Republic of Korea. References [1] J. Olson, S. Nordin, in Duplex Stainless Steels Symp. Proc., The Hague, 1986, pp. 219–225. [2] E. Perteneder, J. Tosch, P. Reiterer, G. Rabensteiner, in Duplex Stainless Steels Symp. Proc., The Hague, 1986, pp. 48–56. [3] J.-O. Nilsson, Mater. Sci. Technol. 8 (1992) 685–700. [4] R.F. Steigerwald, Corrosion 33 (1977) 338–344. [5] H. Kiesheyer, H. Brandis, Z. Metallkd. 67 (1976) 258–263. [6] N. Lopez, M. Cid, M. Puiggali, Corros. Sci. 41 (1999) 1615–1631. [7] J.O. Nilson, P. Kangas, T. Karlsson, A. Wilson, Metall. Mater. Trans A 31 (2000) 35–45. [8] N.O. Lara, A. Ruiz, C. Rubio, R.R. Ambriz, A. Medina, NDT & E Int. 44 (2011) 463– 468. [9] K.H. Lo, J.K.L. Lai, C.H. Shek, D.J. Li, Mater. Sci. Eng. A 452–453 (2007) 149– 160. [10] A.J. Strutt, G.W. Lorimer, C.V. Roscoe, K.J. Gradwell, Proceedings of the Duplex Stainless Steels’86, 1986, p. 310. [11] C.S. Huang, C.C. Shih, Mater. Sci. Eng. A 402 (2005) 66–75. [12] K. Ogawa, H. Okamoto, M. Igarashi, M. Ueda, T. Mori, T. Kobayashi, Welding Int. 11 (1997) 14–22. [13] J.S. Kim, H.S. Kwon, Corrosion 55 (1999) 512–521. [14] C.J. Park, M.K. Ahn, H.S. Kwon, Mater. Sci. Eng. A 418 (2006) 211–217. [15] V.T. Borisov, V.M. Golikov, G.V. Sherbedinskiy, Phys. Met. Metall. 22 (1996) 175. [16] X. Li, A.P. Miodownik, N. Saunders, Mater. Sci. Technol. 18 (2002) 861–868. [17] Annual Book of ASTM Standards, ASTM G 5. [18] M. Pohl, O. Storz, Z. Metallkd. 95 (2004) 631–638. [19] M. Pohl, O. Storz, Mater. Charact. 58 (2007) 65–71. [20] C.J. Park, V.S. Rao, H.S. Kwon, Corrosion 61 (2005) 76–83. [21] K.N. Adhe, V. Kain, K. Madangopal, J. Mater. Eng. Perform. 5 (1996) 500– 506. [22] H.L. Cao, S. Hertzman, W.B. Hutchision, Stainless steels’87, The Institute of Metals, The University of York, September 1988, p. 454. [23] J. Dobranszky, P.J. Szabo, T. Berecz, V. Hrotko, M. Portko, Spectrochim. Acta, Part B 59 (2004) 1781–1788.