Effect of hydrogen on the hardness of different phases in super duplex stainless steel

Effect of hydrogen on the hardness of different phases in super duplex stainless steel

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e9 Available online at www.sciencedirect.com ScienceDire...

2MB Sizes 2 Downloads 30 Views

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e9

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.elsevier.com/locate/he

Effect of hydrogen on the hardness of different phases in super duplex stainless steel Nousha Kheradmand a,*, Roy Johnsen a, Jim Stian Olsen b,c, Afrooz Barnoush a a

Department of Engineering Design and Materials, Richard Birkelands vei 2b, 7491 Trondheim, Norway Department of Structural Engineering, Richard Birkelandsvei 1A, 7491 Trondheim, Norway c Aker Solutions, Fornebu, Norway b

article info

abstract

Article history:

Despite its superior corrosion resistance, super duplex stainless steels (SDSS) are prone to

Received 13 July 2015

hydrogen embrittlement. In this paper, a novel in situ electrochemical nanoindentation

Received in revised form

technique is used to investigate the hydrogen effect on the nanomechanical response of

4 October 2015

the existing phases in SDSS, i.e. ferrite and austenite. A systematic change in electro-

Accepted 26 October 2015

chemical (EC) charging potential revealed the interconnected nature of the hydrogen effect

Available online xxx

on the nanomechanical properties of SDSS. It is shown that the hydrogen effects in each phase are very different and are strongly coupled with the existing residual stresses in the

Keywords:

microstructure induced during the manufacturing and/or induced by EC hydrogen

Super duplex stainless steel

charging.

Austenite

Copyright © 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights

Ferrite

reserved.

Electrochemical nanoindentation Nano hardness Residual stress

Introduction Super duplex stainless steels (SDSS) are two-phase highly alloyed steels with excellent resistance against localized corrosion [1e3]. They are used in especially demanding environments like the oil and gas industry and the chemical industry. However, despite the very favorable properties of SDSS, they tend to fail when exposed to hydrogen, due to hydrogen embrittlement [4e13]. In spite of much research into hydrogen embrittlement (HE) in SDSS, its mechanism in this alloy is still not completely understood [14]. This is mainly

because HE is the result of the interaction of a material under a mechanical load with its environment, as shown in Fig. 1. Numerous circumstances can result in the ingress of hydrogen into the metal. Mechanical loading can manifest itself in different ways, e.g. residual stresses, cyclic loading, static loading, etc. The third aspect, beside the environmental and mechanical ones, addresses the wide range of intrinsic and extrinsic variables within the material itself. In the case of SDSS, we are dealing with a two-phase material comprised of austenite (g) and ferrite (a) with different physical properties. Specifically, from the HE point of view, these differences are crucial, e.g. austenite has a high H

* Corresponding author. E-mail addresses: [email protected] (N. Kheradmand), [email protected] (R. Johnsen), [email protected] (J.S. Olsen), [email protected] (A. Barnoush). http://dx.doi.org/10.1016/j.ijhydene.2015.10.106 0360-3199/Copyright © 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved. Please cite this article in press as: Kheradmand N, et al., Effect of hydrogen on the hardness of different phases in super duplex stainless steel, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.10.106

2

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e9

Fig. 1 e Different aspects of hydrogen embrittlement and complication.

solubility and low diffusivity, while ferrite is vice versa. The glide resistances to dislocations, widely referred to as PeierlseNabarro resistance, which results from the pulsing distortions of the dislocation core as it moves through the discrete lattice, are also different for g austenite and a ferrite [15]. The PeierlseNabarro resistance often affects edge and screw dislocations differently, and at very distinct levels in different crystal structures can also be affected by hydrogen differently. Therefore, it is necessary to study the HE on similar scales in the individual phases. For the purpose of local mechanical testing, instrumented nanoindentation is the most appropriate tool. The measured loadedisplacement (LeD) curves of nanoindentation (Fig. 2) can be analyzed to extract hardness and elastic moduli according to the wellknown OliverePharr method, as given in Eqs (1) and (2) [16]. H¼

Pmax Ac

(1)

Er ¼

pffiffiffi S p pffiffiffiffiffiffi 2b Ac

(2)

Here, S is the slope of the loadedisplacement curve at the initial unloading, as shown in Fig. 2, Ac is the projected contact area evaluated from the contact depth hc shown in Fig. 2 and the tip area function. b is a correction factor dependent on the tip geometry (1.034 for the Berkovich indenter used in this study). In order to probe the hydrogen effect during nanoindentation, the best practice is achieved by combining nanoindentation with in situ hydrogen charging (H-charging) [17e19]. It has been shown that the presence of hydrogen in a metal can alter the nanomechanical footprint registered in the form of the LeD curve [20e25]. Hydrogen can affect the hardness, elastic modulus, and/or the elastic to plastic transition load (pop-in load). The purpose of this paper is to examine and understand the effect of EC H-charging at different conditions separately on the phases present in SDSS.

Experimental

Load, P

Material

Unloading

Pmax

S

Loading

hf

hmax

The SDSS sample was a SAF 2507 provided by OUTOKUMPU. This grade is comparable to the American UNS S32750 and European EN 1.4410 grades. The composition of the sample provided by the supplier is given in Table 1. The sample was in the solution-annealed condition, i.e. heat treated at 1120 for 45 min and quenched in water. The ferrite content of the sample was 46% according to the supplier's test results. The chemical composition of the main alloying elements in the sample was measured by using energy-dispersive X-ray spectroscopy (EDS) and is shown in Table 2. As expected, the ferrite stabilizing elements, chromium and molybdenum, are

hc

Displacement, h Fig. 2 e Typical nanoindentation loading and unloading curve.

Table 1 e Composition of SDSS used in this study provided by the supplier. Element wt.%

C

Si

Mn

P

S

Cr Ni

Mo Cu

N

0.016 0.23 0.79 0.021 0.001 25 6.98 3.82 0.32 0.27

Please cite this article in press as: Kheradmand N, et al., Effect of hydrogen on the hardness of different phases in super duplex stainless steel, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.10.106

3

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e9

Table 2 e Composition of SDSS used in this study measured by EDS. Element (wt. %) Mean Ferrite Austenite

Fe

Cr

Ni

Mo

62.2 61.5 63.5

26 27 24.7

7 5.7 8.2

4.8 5.8 3.7

enriched in the ferrite, and the austenite stabilizing element, nickel, is enriched in the austenite. Standard surface preparation, including mechanical polishing to 1 mm with diamond paste followed by electropolishing in H2 SO4/methanol [26], was used. Because of the different surface binding energies of ferrite and austenite, the connected grains of both phases can be easily distinguished based on their shape. The austenite grain clusters are preferentially convex, whereas the ferrite grains fill the space between these grain clusters (Fig. 3).

In situ electrochemical nanoindentation (ECNI) A Hysitron TI 950 TriboIndenter™ with a Performech™ control unit nanoindentation system together with a threeelectrode EC setup with a platinum counter electrode was used for the ECNI measurements. All of the EC potentials in this paper are reported against the Hg/HgSO4 (Mercury/ Mercurous Sulfate) reference electrode. A Gammery Reference 600™ Potentiostat/Galvanostat was used to control the EC potentials or current density, and the EC data were recorded on a computer. Further details of the experimental setup are given elsewhere [27e29]. All indents were made using a Berkovich diamond tip, which was also used to image the surface of the sample prior to and after indentation with the in-situ imaging option on the Nanoindentation system. The resulting surface topography images were further analyzed using open source Gwyddion scanning probe microscope analysis software [30] to measure the surface roughness of the nanoindented regions for each polarization potential. Almost all of the LeD curves showed a pop-in, however, in this paper we will focus only on the effect of

EC hydrogen charging on the hardness. In order to avoid the effect of pop-in load variations on the hardness, we specifically pay attention to when there is a maximum load of indentation well above the pop-in load. This ensured that the variations of the pop-in load would not affect the hardness measurements. The roughness analysis was performed over a small area of 1.6 mm2, comparable with the typical indent size. The hardness and elastic modulus values were calculated from the LeD curves according to the OliverePharr method [31]. Indents on fused quartz were used to calibrate the tip area function. The effect of in situ EC hydrogen charging was studied in a solution of 2:1 glycerol and phosphoric acid [32] in combination with potentiostatic Hcharging. Different EC potentials result in different subsurface hydrogen concentrations. In order to find the dependence of the nanomechanical response of each existing phase in the SDSS on the hydrogen concentration, cathodic and anodic potentials were applied to the sample surface. Several indentation sets were performed in situ in the electrolyte under cathodic and anodic potentials, but the first set of indentations was performed in the air as the reference point. Under a cathodic potential (CP), the SDSS sample surface gets negatively charged and positive hydrogen ions are reduced to atomic hydrogen on the surface. While part of the atomic hydrogen are combined to form H2 molecules, part of this atomic hydrogen ingresses into the metal and dissolves in it. Under an anodic potential (AP), the surface gets positively charged, and dissolved hydrogen atoms in the metal oxidize on the surface so that the surface discharges hydrogen. For evaluation of the hydrogen effect, the data obtained at different potentials are always compared in the same grain. At least 10 indents in each sequence of potentials were performed.

Results Austenite Typical LeD curves during nanoindentation for the austenite phase in SDSS are shown in Fig. 4. While the LeD curves

600

Anodic (100 mV)

Load (μN)

500 Cathodic (-925 mV)

400 300

Air

200 100

Fig. 3 e Differential interference contrast optical microscope image of the electropolished SDSS sample. The connected grains of both austenite and ferrite phases can be easily distinguished based on their shape. The austenite grain clusters are preferentially convex, whereas the ferrite grains fill the space between these grain clusters.

0

0

10

20

30

40

Displacement (nm)

Fig. 4 e Representative load displacement curves during nanoindentation of the austenite phase in SDSS under anodic (100 mV) and cathodic (¡925 mV) conditions.

Please cite this article in press as: Kheradmand N, et al., Effect of hydrogen on the hardness of different phases in super duplex stainless steel, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.10.106

4

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e9

Fig. 5 e Variation of nanohardness of the austenite phase by changing the in situ electrochemically charged hydrogen in SDSS.

measured in air and under AP at 100 mV are very similar, there is a clear change in the LeD curves after application of CP at 925 mV. Therefore, special attention was paid to the variation of the hardness in the austenite at different applied EC potentials.

The variation of nanohardness in the austenite phase with EC potential is shown in Fig. 5. Each point presented in Fig. 5 is representative of the mean value of the hardness under the given applied EC potential. The type of the potential (AP/CP) followed by its implementation sequence number are given for each point. The error bars are the standard deviations. Clearly, by application of the first sequence of the CP of 900 mV (CP1) the mean hardness from 5 GPa in air increased to 5.4 GPa. Following AP at 100 mV (AP2), the hardness went back to something close to its hardness in air. Further sequences of cathodic and anodic charging show the same behavior, i.e. hardness increases with the applied CP. By decreasing the CP, in other words, by increasing the absolute value of the CP, the hardness increases. Surface topography images of the tested austenite grain are shown in Fig. 6. The measured arithmetic average roughness (Ra) and the root mean square roughness (RRMS) under different conditions are presented in Table 3. The measured roughness values show very small changes under the different conditions, which ensures that the observed changes in the nanomechanical behavior is not due to the changes in the surface roughness. A high resolution plot of an indent made in Austenite at CP6 (Fig. 6c) and the piled-up material around the indents is shown in Fig. 7.

Fig. 6 e Topography images of the austenite surface in (a) Air, (b) CP1 (¡900 mV), the new indents made in CP1 are marked with circles, (c) CP6 (¡925 mV) and (d) AP7 (100 mV), indents in solid line circle are the new indents at AP7 and the ones in dashed line circle are previous indents at CP6. Please cite this article in press as: Kheradmand N, et al., Effect of hydrogen on the hardness of different phases in super duplex stainless steel, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.10.106

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e9

5

Table 3 e Arithmetic average roughness (Ra) and root mean square roughness (RRMS) values at different conditions measured from the topography images. Austenite Condition Air CP1 AP2 CP6 AP7

Ra (nm) 0.11 0.16 e 0.21 0.15

Ferrite

RRMS (nm) 0.14 0.20 e 0.26 0.19

Ra (nm) 0.13 0.13 0.13 0.2 0.22

RRMS (nm) 0.16 0.16 0.17 0.29 0.27

Fig. 9 e Effect of in situ electrochemically-charged hydrogen on nanohardness of the ferrite phase in SDSS.

Fig. 7 e 2D color filled contour plot of one of the indents made in Austenite at CP6 (Fig. 6(c)).

Ferrite The LeD curves during nanoindentation on the ferrite phase in SDSS are presented in Fig. 8, where obvious changes in the LeD curves with the different conditions of AP at 100 mV and CP at 900 mV can be seen. Analyzing the nanoindentation data of the ferrite phase in different conditions resulted in the hardness values shown in Fig. 9. In comparison to the measured hardness values for austenite under different conditions, the results for the ferrite 600

Cathodic (-900 mV)

500

Load (μN)

400

Air

300 200

100

Anodic (100 mV)

0 0

10

20

30

40

Displacement (nm)

Fig. 8 e Representative load displacement curves during nanoindentation of the ferrite phase in SDSS at anodic (100 mV) and cathodic (¡900 mV) conditions.

phase are more complex. In the first sequence of CP at 900 mV (CP1), there is almost no change observed in the hardness of the ferrite. But, after application of 100 mV (AP2), the hardness significantly reduced. Following this reduction in the hardness, subsequent hydrogen charging in the next CP steps, CP5 to CP9, always resulted in an increase in the hardness. This increase in the hardness correlates with the applied potential and increases as the negative CP increases. After AP4 and AP7, the hardness returns back to the measured value at the first anodic charging at AP2. Surface topography images of the tested ferrite grain are shown in Fig. 10 and the roughness values at different conditions are presented in Table 3. As in the case of the austenite phase, the measured roughness values again show very little changes under the different conditions, which ensures that the observed changes in the nanomechanical behavior are not due to the changes in the surface roughness.

Discussion Hydrogen solubility in SDSS Interpretating the variations in the nanohardness in the austenite and ferrite phases in SDSS under different conditions, as shown in the previous section, first of all requires a clear understanding of the complex interaction of these phases during the manufacturing process. Typically, for SDSS a microstructure composed of 50% ferrite and 50% austenite is desired. Such a microstructure is obtained by quenching the alloy from about 1120 to room temperature in order to stabilize the metastable austenite phase. When the multi-phase alloy is heat-treated, internal stresses form, due to differences in the thermal expansion coefficients of the constituent phases. The distribution of such internal stresses in SDSS after quenching in water from an elevated temperature has been discussed in the literature [33e35]. These stresses can be divided into macroscopic and microscopic residual stresses. The macroscopic residual stress, sM ij , is due to the temperature gradient between the surface and the internal region of the specimen. If the

Please cite this article in press as: Kheradmand N, et al., Effect of hydrogen on the hardness of different phases in super duplex stainless steel, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.10.106

6

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e9

Fig. 10 e Topography images of the ferrite surface in (a) Air, (b) CP1 (¡900 mV), indents marked with solid line are made at CP1 and previous indents made in air are marked with dashed line, (c) AP7 (100 mV), indents made in AP7 are marked with solid line and previous indent from CP1 are marked with dashed line. The unmarked indents here are those made in air.

elastic moduli of the constituent phases are not too different, as in the case of SDSS, then sM ij can be neglected [33]. The microscopic residual stresses, however, arise due to the misfit strains among the grains of the constituent phases with different thermal expansion coefficients. These stresses may vary from grain to grain, and their averaged value over each constituent ph phase is called the thermal phase-stress, sij , [35]. We used Eshelby's inclusion theory coupled with the MorieTanaka mean field theory [36] to calculate the average stress in the matrix, i.e. ferrite, resulting in the value sFij ¼ 470 MPa (compressive) and in the randomly distributed inclusions, i.e. austenite, which resulted in sA ij ¼ 400 MPa (tensile) [9]. These values are in good agreement with the reported values in the literature calculated with the constitutive material law (~400 MPa) [37], measured with neutron diffraction (~500 MPa) [33] and x-ray diffraction (~280e500 MPa) [35]. This relatively high residual stress plays a crucial role during the hydrogen charging of SDSS at CPs. Primarily, it increases the solubility of hydrogen in austenite according to the following equation [38]:   sVH Cs ¼ C0 exp RT

(3)

in which Cs and C0 are the concentration of interstitial solute in the stressed and unstressed body, respectively, and VH is the partial molal volume of the hydrogen with a typical value of VH z 2 cm3 mol1. For tensile stress, s is negative, and so the ph solubility increases. The calculated sij for the two phases means an approximately 38% increase in solubility of hydrogen in austenite and ~32% decrease in solubility of hydrogen in ferrite. Unfortunately, no information on the solubility and diffusivity of H in each of the present phases in

SDSS is available. This is due to the unavailability of a reference alloy, completely austenitic or ferritic with the identical chemical composition as existing in SDSS phases. However, there are austenitic or ferritic steels available, having relatively similar compositions. For example, AL 29-4-2 (UNS S44800), which has about 29% Cr, 4%Mo and 2%Ni, and the 21Cre6Nie9Mn (UNS S21900/S21904) austenitic stainless steel are comparable with the ferrite and austenite phases of the SDSS respectively (see Table 2). The H permeability and diffusivity in the 21Cre6Nie9Mn austenitic stainless steel [39] and in AL 29-4-2 ferritic stainless steel has been studied [40]. Both the diffusivity, D, and permeability, F, are thermally activated processes. Thus, they feature an Arrhenius-type dependence on temperature an d so can be expressed as   DHD D ¼ D0 exp RT

(4)

  DHF F ¼ F0 exp RT

(5)

respectively, where D0 and F0 are temperature independent preexponentials, DHD and DHF are the corresponding activation energies, R is the gas constant and T the temperature. Since F ≡ DS, the solubility S can be determined from the ratio of F and D, and by inserting Eqs (4) and (5) [39], there results the following relation S¼

  F0 ðDHF  DHD Þ exp RT D0

(6)

Table 4 summarizes the calculated solubilities from the diffusivity and permeability data according to Eq. (6). As

Table 4 e Solubility of hydrogen in different stainless steels with comparable composition to the existing phases in the SDSS. pffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffi Steel S0 ðmolH2 =m3 MPaÞ Sð298 Þ ðmolH2 =m3 MPaÞ DHFDHD Ref. AL 29-4-2 (Ferr) 21Cre6Nie9Mn (Aus) 300-Series (Aus)

1534 222 135

31.4 5.9 5.9

0.005 20.52 12.48

[40] [39] [39]

Please cite this article in press as: Kheradmand N, et al., Effect of hydrogen on the hardness of different phases in super duplex stainless steel, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.10.106

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e9

expected, the hydrogen solubility in ferrite is pffiffiffiffiffiffiffiffiffiffi 0:005 molH2 =m3 MPa, while in austenite, depending on the composition of the austenite phase, between 12.48 and pffiffiffiffiffiffiffiffiffiffi 20:52 molH2 =m3 MPa are reported. According to the Sieverts law, the concentration of hydrogen dissolved in the metal lattice in equilibrium with the hydrogen qffiffiffiffiffiffigas, can be expressed as [32,39]: (7) cH ¼ S fH2 where fH2 is the hydrogen fugacity. The equivalent fugacity of the hydrogen during EC charging depends on the hydrogen evolution reaction. In the case of ferrous alloys, we can assume a fast discharge reaction, and then the following relation applies [41]:   2hF fH2 ¼ exp RT

(8)

where h ¼ EE0 is the overpotential, E is the applied potential, E0 is the equilibrium reduction potential, and F is the Faraday constant. Inserting Eqs (3) and (8) into (7) results in sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi    2hF sVH exp Cs ¼ S exp RT RT

(9)

which relates the applied EC overpotential h to the H concentration in a stressed lattice CH,s. Assuming that the hydrogen evolution mechanism on the surfaces of both austenite and ferrite phases in SDSS are identical, Eq. (9) shows that during each potential sequence, the subsurface hydrogen concentration depends only on the solubility of hydrogen and the amount of the residual stress in the individual phases. Thus, a hydrogen partitioning coefficient in austenite and ferrite, PHAus=Ferr , can be defined: PHAus=Ferr ¼

  F ðs  sA ÞVH SA exp SF RT

(10)

Using the typical values given in Table 4 and the determined residual stresses, we calculated the hydrogen partitioning coefficient of PHAus=Ferr ¼ 8330. This means that in SDSS, the hydrogen concentration at any CP in the austenite is expected to be 8330 times higher than in the ferrite. Discussing any experimental observation of the effect of hydrogen on the mechanical properties of SDSS should consider this difference in the solubility of the two existing phases.

The effect of hydrogen on the nanohardness of SDSS The effect of residual stresses on nanoindentation behavior has been extensively discussed in the literature [42e47]. It has been shown that the indentation parameters, such as the curvature of the loading curve, the elastically recovered depth (hmaxhf in Fig. 2), the residual depth (hmax in Fig. 2), the indentation work (the area within the loading and unloading curves in Fig. 2), and the pile-up amount (red colored area (in the web version) in Fig. 7), are all affected by the presence of residual stresses. Compared to the stress-free state, compressive stress increases the curvature of loading curve, the elastically recovered depth, the pile-up height, and the contact area, whereas it decreases the residual depth [46]. Similarly, the effect of tensile stress on these parameters is

7

the opposite to that of the compressive stress state. Therefore, the measured hardnesses in both the ferrite and austenite phases are affected by the presence of residual stresses. In the case of ferrite, due to the presence of the compressive stress state, the hardness is higher than the actual hardness of the ferrite. On the other hand, in the austenite, due to the presence of tensile stresses, the measured hardness is expected to be lower than the actual real value of the stress free austenite. Unfortunately, since there is no ferritic or austenitic steel with the typical composition of the present phases in the SDSS existing, there is no possibility of evaluating the hardness values of stress free austenite and ferrite. Attempts to perform a heat treatment to relieve the stress results in the formation of nanoscale precipitates like Cr2N, which affect the hardness [48]. The comparable AL 29-4-2 (UNS S44800) super ferritic and the 21Cre6Nie9Mn (UNS S21900/S21904) austenitic stainless steels are reported by different producers to have a similar macroscopic hardness, close to HRB 100 (~2.5 GPa). However, the macroscopic hardness is not comparable with the nanohardness, which includes a size effect and the depth dependence of the hardness [49]. Additionally, the effect of interstitials like nitrogen and carbon on the hardness is more than the substitutional elements. Therefore, a comparison with the above mentioned alloys and any conclusions on the basis of this comparison should be made very carefully. The high solubility of hydrogen in austenite, which is in turn enhanced by the presence of the residual tensile stress, results in the formation of compressive stresses [50,23,51] and is expected to neutralize the existing tensile thermal phasestress discussed in Section 4.1. We believe that this relaxation of the thermal phase stress by hydrogen-induced compressive stress in austenite in turn results in a relaxation of the compressive stress in the neighboring ferrite phase. It manifests itself as an observed unusual softening and irreversible relaxation of residual stress in the ferrite after the first cycle of hydrogen charging discharging, shown in Fig. 9. In contrast to ferrite, no increase in the hardness of the austenite due to the relaxation of the tensile stress was observed after the first cycle of hydrogen charging discharging (CP1 followed by AP2). Considering the trapping of the hydrogen in the already strained lattice of the austenite, the application of AP is not high enough to extract this trapped and thermodynamically stable hydrogen from out of the lattice. The hardness of the ferrite phase gradually increases when increasing the CP, i.e. the hydrogen content in the ferrite. The source of this increase in the hardness can be both the presence of the hydrogen in the lattice and its interaction with the moving dislocations, i.e. the Cottrel cloud [24], or the compressive stress exposed by the expansion of the austenite grains by hydrogen uptake. Previous research on the effect of hydrogen on the hardness of ferritic alloys has been controversial. Zhao et al. showed that, depending on the charging condition, both hardening and softening can be observed, though EC hydrogen charging resulted in an increase in the hardness of low carbon steel in their paper [52]. Lee et al. [53] showed a change from hardening to softening by hydrogen, depending on the geometry of the tip and its representative strain. Finally, theoretical studies of Itakura et al. [54] showed that, depending on the hydrogen concentration and temperature, softening or hardening can be expected in iron in the presence of the hydrogen.

Please cite this article in press as: Kheradmand N, et al., Effect of hydrogen on the hardness of different phases in super duplex stainless steel, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.10.106

8

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e9

The hardness of the austenite also increases when increasing the hydrogen concentration in austenite by increasing the CP. The increase in the hardness of the austenite as proposed by Takakuwa et al. [23] is due to a mixture of hydrogen dislocation interaction and formation of compressive residual stresses. However, the variation of the hardness is not as uniform as in the case of the ferrite. The hardness shows a sudden increase at 925 mV (CP6). Interestingly, the surface images of the indents made at this potential show a different topography, as shown in Fig. 6(c). Clearly, the formation of surface features after nanoindentation in austenite at potentials greater than or equal to 925 mV (CP6) are observable. In addition, application of an AP is enough to remove the hydrogen and recover the material behavior, i.e. inhibit the formation of these surface features, as is shown in Fig. 6(d) (the new indents within the solid line). An exact explanation for the formation of these features is still missing. It might be due to the formation of ε or a0 martensite as a result of the reduction in stacking fault energy by hydrogen, as reported by Rozenak [50,55].

Conclusion The effect electrochemically charged hydrogen on both the austenite and ferrite phases in SDSS was studied by means of the in situ ECNI technique. The near surface hydrogen concentration was varied within the two phases through the application of different CPs. It has been shown that the electrochemically charged hydrogen increases the hardness of the two phases in SDSS. The increase in the hardness scaled with the applied potential shows the scaling of the hardness with the hydrogen concentration, as can be inferred from Eq. (9). The return of the hardness to its original value after EC oxidation of the dissolved hydrogen at AP also confirms that the observed increase in the hardness is due to the presence of hydrogen in the lattice. The presented nanomechanical examination of SDSS reveals the importance of the interplay between the residual stress and the hydrogen, and their effect on the mechanical properties. Therefore, for the application of SDSS in environments where it is prone to hydrogen uptake and hydrogen embrittlement, it is necessary to consider the presence of residual stress in the microstructure and its effect on the hydrogen solubility in the different phases. Lastly, the changes in the microstructure and the residual stress during the hydrogen uptake should be taken into account for design purposes.

Acknowledgments The Research Council of Norway is acknowledged for support for the NTNU Nanolab through the Norwegian Micro- and Nano-Fabrication, Norfab (197411/V30). This research was also partly supported by the Research Council of Norway through the project HIPP (Hydrogen-Induced degradation

of offshore steels in aging infrastructuredModels for Prevention and Prediction) under the project number 234130/E30.

references

[1] Turnbull A, Francis PE, Ryan MP, Orkney LP, Griffiths AJ, Hawkins B. A novel approach to characterizing the corrosion resistance of super duplex stainless steel welds. Corrosion 2002;58(12):1039e48. [2] Hoseinpoor M, Momeni M, Moayed M, Davoodi A. EIS assessment of critical pitting temperature of 2205 duplex stainless steel in acidified ferric chloride solution. Corros Sci 2014;80(0):197e204. [3] Lauvstad G, Johnsen R, Asbjrnsen I, Bjurstrm M, Hjorth CG. Resistance toward hydrogen-induced stress cracking of hot isostatically pressed duplex stainless steels under cathodic protection. Corrosion 2010;66(11):115004e13. [4] Oltra R, Bouillot C, Magnin T. Localized hydrogen cracking in the austenitic phase of a duplex stainless steel. Scr Mater 1996;35(9):1101e5. [5] Takakuwa O, Soyama H. Suppression of hydrogen-assisted fatigue crack growth in austenitic stainless steel by cavitation peening. Int J Hydrogen Energ 2012;37(6):5268e76. [6] Zakroczymski T, Owczarek E. Electrochemical investigation of hydrogen absorption in a duplex stainless steel. Acta Mater 2002;50(10):2701e13. [7] El-Yazgi A, Hardie D. Stress corrosion cracking of duplex and super duplex stainless steels in sour environments. Corros Sci 1998;40(6):909e30. [8] Borchers C, Michler T, Pundt A. Effect of hydrogen on the mechanical properties of stainless steels. Adv Eng Mater 2008;10:11e23. [9] Barnoush A, Zamanzade M, Vehoff H. Direct observation of hydrogen-enhanced plasticity in super duplex stainless steel by means of in situ electrochemical methods. Scr Mater 2010;62:242e5. [10] Olden V, Thaulow C, Berstad T, Østby E, Nyhus B. Influence of hydrogen from CP on the fracture susceptibility of 25% Cr duplex stainless steel e FE modeling of constant load testing using hydrogen influenced cohesive zone elements. 2008. p. 452e8. [11] Olden V, Thaulow C, Johnsen R. Modelling of hydrogen diffusion and hydrogen induced cracking in supermartensitic and duplex stainless steels. Mater Des 2008;29(10):1934e48. [12] Olden V, Thaulow C, Johnsen R, Østby E, Berstad T. Application of hydrogen influenced cohesive laws in the prediction of hydrogen induced stress cracking in 25% Cr duplex stainless steel. Eng Fract Mech 2008c;75:2333e51. [13] da Silva B, Salvio F, dos Santos D. Hydrogen induced stress cracking in UNS S32750 super duplex stainless steel tube weld joint. Int J Hydrogen Energy 2015;40(47):17097e101. [14] Lo K, Shek C, Lai J. Recent developments in stainless steels. Mater Sci Eng R 2009;65(4e6):39e104. [15] Argon AS. Strengthening mechanisms in crystal plasticity. Oxford University Press Oxford; 2008. [16] Oliver W, Pharr G. An improved technique for determining hardness and elastic modulus using load and displacement sensing indentation experiments. J Mater Res 1992;7(6):1564e83. [17] Barnoush A, Bies C, Vehoff H. In situ electrochemical nanoindentation of FeAl (100) single crystal: hydrogen effect on dislocation nucleation. J Mater Res 2009;24(3):1105e13. [18] Barnoush A, Vehoff H. Hydrogen embrittlement of aluminum in aqueous environments examined by in situ electrochemical nanoindentation. Scr Mater 2008;58:747e50.

Please cite this article in press as: Kheradmand N, et al., Effect of hydrogen on the hardness of different phases in super duplex stainless steel, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.10.106

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e9

[19] Barnoush A, Vehoff H. In situ electrochemical nanoindentation of a nickel (111) single crystal: hydrogen effect on pop-in behaviour. Int J Mat Res (formerly Z Metallkd) 2006;97:1224e9. [20] Mine Y, Doi K, Matsuoka S, Murakami Y. The influences of hydrogen on microscopic plastic deformation behavior of SUS304 and SUS316l stainless steels. Zairyo J Soc Mater Sci Jpn 2008;57(3):255e61. [21] Barnoush A, Vehoff H. Recent developments in the study of hydrogen embrittlement: hydrogen effect on dislocation nucleation. Acta Mater 2010;58(16):5274e85. [22] Nibur K, Bahr D, Somerday B. Hydrogen effects on dislocation activity in austenitic stainless steel. Acta Mater 2006;54:2677e84. [23] Takakuwa O, Mano Y, Soyama H. Increase in the local yield stress near surface of austenitic stainless steel due to invasion by hydrogen. Int J Hydrogen Energ 2014;39(11):6095e103. [24] Barnoush A, Asgari M, Johnsen R. Resolving the hydrogen effect on dislocation nucleation and mobility by electrochemical nanoindentation. Scr Mater 2012;66:414e7. [25] Zhang L, An B, Fukuyama S, Yokogawa K. Hydrogen effects on localized plasticity in SUS310S stainless steel investigated by nanoindentation and atomic force microscopy. Jpn J Appl Phys 2009;48:08JB08. [26] Asgari M, Barnoush A, Johnsen R, Hoel R. Microstructural characterization of pulsed plasma nitrided 316L stainless steel. Mater Sci Eng A 2011;529:425e34. [27] Asgari M, Johnsen R, Barnoush A. Nanomechanical characterization of the hydrogen effect on pulsed plasma nitrided super duplex stainless steel. Int J Hydrogen Energy 2013;38:15520e31. [28] Barnoush A, Basa A, Thaulow C. Oxygen argon plasma treatment effect on hydrogen uptake in austenitic stainless steels. Int J Hydrogen Energ 2014;39(26):14120e31. [29] Zamanzade M, Vehoff H, Barnoush A. Cr effect on hydrogen embrittlement of iron aluminide intermetallics: surface or bulk effect. Acta Mater 2014;69:210e23. as D, Klapetek P. Gwyddion: an open-source software for [30] Nec SPM data analysis. Central Eur J Phys 2012;10:181e8. [31] Oliver W, Pharr G. Measurement of hardness and elastic modulus by instrumented indentation: advances in understanding and refinements to methodology. J Mater Res 2004;19(1):3e20. [32] Gemma R, Dobron P, Cizek J, Pundt A. Stress release and defect occurrence in V1x Fex films upon hydrogen loading: H-induced superabundant vacancies, movement and creation of dislocations. Acta Mater 2014;67(0):308e23. [33] Harjo S, Tomota Y, Ono M. Measurements of thermal residual elastic strains in ferrite-austenite fe-cr-ni alloys by neutron and x-ray diffractions. Acta Mater 1998;47(1):353e62. [34] Pohl M, Bracke A. Thermally induced microstructural changes of ferritic-austenitic stainless steels and their influence on residual stress. Z Metallkd Mater Res Adv Tech 1999;90(8):551e6. n M, Zeng X. Evolution of the residual stress [35] Johansson J, Ode state in a duplex stainless steel during loading. Acta Mater 1999;47(9):2669e84. [36] Mura T. Micromechanics of defects in solids. Kluwer Acad Pub 1987;177. [37] Siegmund T, Werner E, Fischer FD. On the thermomechanical deformation behavior of duplex-type materials. J Mech Phys Solids 1995;43(4):495e501.

9

[38] Bockris J. Modern electrochemistry: an introduction to an interdisciplinary area. Plenum Press; 2000. [39] Marchi CS, Somerday B, Robinson S. Permeability, solubility and diffusivity of hydrogen isotopes in stainless steels at high gas pressures. Int J Hydrogen Energy 2007;32(1):100e16. [40] Tsong-Pyng P, Altstetter C. Effects of deformation on hydrogen permeation in austenitic stainless steels. Acta Metall 1986;34(9):1771e81. [41] Bockris JO, Subramanyan PK. The equivalent pressure of molecular hydrogen in cavities within metals in terms of the overpotential developed during the evolution of hydrogen. Electrochim Acta 1971;16:2169e79. [42] Zhu LN, Xu BS, Wang HD, Wang CB. Effect of residual stress on the nanoindentation response of (100) copper single crystal. Mater Chem Phys 2012;136:561e5. [43] Khan M, Fitzpatrick M, Hainsworth S, Evans A, Edwards L. Application of synchrotron x-ray diffraction and nanoindentation for the determination of residual stress fields around scratches. Acta Mater 2011;59(20):7508e20. [44] Olivas E, Swadener J, Shen Y. Nanoindentation measurement of surface residual stresses in particle-reinforced metal matrix composites. Scr Mater 2006;54(2):263e8. [45] Carlsson S, Larsson PL. On the determination of residual stress and strain fields by sharp indentation testing: part ii: experimental investigation. Acta Mater 2001;49(12):2193e203. [46] Swadener J, Taljat B, Pharr G. Measurement of residual stress by load and depth sensing indentation with spherical indenters. J Mater Res 2001;16(7):2092. [47] Suresh S, Giannakopoulos A. A new method for estimating residual stresses by instrumented sharp indentation. Acta Mater 1998;46(16):5755e67. [48] Horvath W, Prantl W, Stroisznigg H, Werner E. Microhardness and microstructure of austenite and ferrite in nitrogen alloyed duplex steels between 20 and 500 c. Mater Sci Eng A Struct Mater Prop Microstruct Process 1998;256(1):227e36. [49] Barnoush A. Correlation between dislocation density and nanomechanical response during nanoindentation. Acta Mater 2012;60(3):1268e77. [50] Rozenak P. Stress induce martensitic transformations in hydrogen embrittlement of austenitic stainless steels. Metall Mater Trans A 2014;45(1):162e78. [51] Hoelzel M, Danilkin S, Ehrenberg H, Toebbens D, Udovic T, Fuess H, et al. Effects of high-pressure hydrogen charging on the structure of austenitic stainless steels. Mater Sci Eng A 2004;384(1e2):255e61. [52] Zhao Y, Seok MY, Choi IC, Lee YH, Park SJ, Ramamurty U, et al. The role of hydrogen in hardening/softening steel: influence of the charging process. Scr Mater 2015;107:46e9. [53] Lee DH, Lee JA, Seok MY, Baek UB, Nahm SH, Jang JI. Stressdependent hardening-to-softening transition of hydrogen effects in nanoindentation of a linepipe steel. Int J Hydrogen Energy 2014;39(4):1897e902. [54] Itakura M, Kaburaki H, Yamaguchi M, Okita T. The effect of hydrogen atoms on the screw dislocation mobility in bcc iron: a first-principles study. Acta Mater 2013;61(18):6857e67. [55] Rozenak P, Shani E. Analysis of phase distribution in thin surface layers comparable to the penetration depth of x-rays. Metall Mater Trans A 2012;43(11):4028e42.

Please cite this article in press as: Kheradmand N, et al., Effect of hydrogen on the hardness of different phases in super duplex stainless steel, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.10.106