- Email: [email protected]
Precipitation of sigma and chi phases in δ-ferrite of Type 316FR weld metals
M A TE RI A L S CH A RACT ER IZ A TI O N 86 (2 0 1 3 ) 1 5 2–1 6 6
Available online at www.sciencedirect.com
ScienceDirect www.elsevier.com/locate/matchar
Precipitation of sigma and chi phases in δ-ferrite of Type 316FR weld metals Eun Joon Chuna,⁎, Hayato Babaa , Kazutoshi Nishimotob , Kazuyoshi Saidaa a Division of Materials and Manufacturing Science, Graduate School of Engineering, Osaka University, 2-1 Yamada-oka, Suita-shi, Osaka 565-0871, Japan b Department of the Application of Nuclear Technology, Fukui University of Technology, Gakuen 3-6-1, Fukui-shi, Fukui 910-8505, Japan
AR TIC LE D ATA
ABSTR ACT
Article history:
The decomposition behavior and kinetics of δ-ferrite are examined using aging treatments
Received 23 July 2013
between 873 and 1073 K for Type 316FR stainless steel weld metals with different
Received in revised form 25
solidification modes (316FR AF, 316FR FA). The dominant precipitates are sigma, chi, and
September 2013
secondary austenite nucleated at δ-ferrite/austenite interfaces or in the interior of the
Accepted 2 October 2013
ferrite grains. These precipitates consume all the ferrite during isothermal aging in both 316FR AF and FA weld metals. Differences in the precipitation behavior (precipitation initiation time and precipitation speed) between weld metals can be explained by і) the
Keywords:
degree of Cr and Mo microsegregation within δ-ferrite or austenite near ferrite and іі) the
Type 316FR
nucleation sites induced due to the solidification mode (AF or FA), such as the ferrite
Weld metal
amount. For both weld materials, a Johnson–Mehl-type equation can express the pre-
Delta ferrite decomposition
cipitation behavior of the sigma + chi phases and quantitatively predict the behavior at the
Sigma phase precipitation
service-exposure temperatures of a fast breed reactor.
Kinetics
© 2013 Elsevier Inc. All rights reserved.
Prediction
1.
Introduction
Compared to common austenitic stainless steels, an advanced Type 316FR austenitic stainless steel has an improved creep rupture resistance and is a key component in the nextgeneration of commercialized fast breed reactors (FBRs) [1–6]. Because a FBR typically operates for about 60 years at high temperatures (773–823 K) due to its outstanding ability to generate heat compared to other types of nuclear reactors (ex. light water-cooled reactor), aging of the structural materials is an inevitable issue. Thus, before FBRs can be commercialized, the aging behavior and the appropriate repair processes must be clarified [7]. One of the most important factors to determine the aged degree of components made of austenitic stainless steels is the inferior microstructural stability of the weld metal at ele-
⁎ Corresponding author. Tel./fax: + 81 6 6879 7543. E-mail address: [email protected] (E.J. Chun). 1044-5803/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.matchar.2013.10.003
vated temperatures. Weld metals of austenitic stainless steel generally involve some amount of δ-ferrite, which is used to prevent solidification cracking during welding. Weld metals are typically composed of different solidification modes, such as primary austenite (γ) with a secondary phase δ-ferrite solidification mode (AF mode) or primary δ-ferrite with a secondary phase γ solidification mode (FA mode) according to the Creq/Nieq variation. However, δ-ferrite may decompose into intermetallic phases (sigma or chi) in a high temperature environment, negatively impacting the mechanical and chemical properties of a weld joint [8]. Hot cracking is a common problem while welding austenitic stainless steel in the AF mode [9]. Because changing the solidification mode or optimizing the δ-ferrite content during solidification can control the hot cracking susceptibility [9], the
153
M A TE RI A L S C HA RACT ER I ZA TI O N 86 ( 20 1 3 ) 1 5 2–1 6 6
chemical composition of the consumable welding components can be adjusted in the FA mode. However excessively increasing the amount of ferrite intensifies the embrittlement of the weld metal because a large amount of intermetallics is transformed while in actual service. Although numerous studies have investigated the aging behavior (precipitation) in austenitic and duplex stainless steels [10–38], information about the kinetics for the weld metal is insufficient compared to that for the base metal [32–38]. Most studies have approximated the kinetics for δ-ferrite decomposition without employing a detailed approach to evaluate individual precipitates [32–38]. Moreover, comparison studies on the precipitation kinetics in weld metals between the AF and FA modes with similar composition ranges are insufficient, despite the possibility that the kinetics may depend on microsegregation according to solidification mode variations. The present study aims to clarify the δ-ferrite decomposition behaviors in Type 316FR stainless steel weld metals utilizing welding consumables with two different solidification modes (AF and FA modes). The decomposition behaviors of ferrite are quantified by accurately measuring the amount of each precipitate. Based upon quantitative analyses, kinetic approaches are pursued and different precipitation behaviors between welding consumables are discussed.
2.
Materials and Experimental Procedure
The materials used are within the compositional range of Type 316FR stainless steel. Table 1 lists their chemical compositions. The designed steel with the AF mode is referred to as 316FR AF, while that with the FA mode is referred to as 316FR FA. The base metal plates of these steels, which measured 120 mm × 40 mm × 3 mm, were manufactured via hot-rolling. To prepare the weld metal, bead-on-plate welding was performed using gas tungsten arc welding (GTAW) on the manufactured plates with an arc current of 110 A, arc voltage of 14 V, and a welding speed of 1.67 mm/s. Welded specimens were aged at various conditions (Table 2) and subsequently quenched with water. The as-welded microstructures were observed by electron backscatter diffraction (EBSD) after electrolytic etching with a 10% aqueous solution of KOH. To investigate the elemental distribution in the as-welded state, an electron probe X-ray micro analyzer (EMPA) measured the concentration profiles of the elements. Precipitates were identified by transmission electron microscopy (TEM) under an acceleration voltage of 200 kV after jet polishing with a solution of perchloric acid (5%) and acetic acid (95%) using an applied voltage of 50 V. To clarify the microstructural changes due to the aging treatment, specimens were observed by scanning electron
Table 2 – Aging conditions. Temperature (K) 873 923 973 1023 1073
Aging time (h) 0.5 0.5 0.5 0.5 0.5
1 1 1 1 1
5 5 5 5 5
10 10 10 10 10
50 50 50 50 50
100 100 100 100 100
500 500 500 500 –
1000 1000 1000 1000 –
microscopy (SEM) and EBSD using both electrolytic etching and mechanical polishing with colloidal silica. Image processing at a 1000-time magnification on the basis of the SEM micrographs quantitatively measured the area fraction of the precipitates.
3. Characterization of Weld Microstructures and Solidification Modes Fig. 1, which shows the as-welded microstructure of the 316FR AF weld metals, confirms the cellular morphology associated with the solidification behavior. All the δ-ferrite is located at the cell boundaries or triple points of γ with an elongated or globular shape. The average volume fraction of ferrite is 2%. These characteristics are consistent with the AF mode [39]. Fig. 2 shows the microstructure of 316FR FA after GTAW bead-on-plate. The average volume fraction of δ-ferrite is 5%. The shape associated with the bands of vermicular δ-ferrite enclosed by γ at the center of a dendritic cell is consistent with the FA mode [40]. Fig. 1 also shows the corresponding pole figures to examine the crystallographic orientation relationship between γ and δ-ferrite. The preferred orientation relationship between γ and δ-ferrite is (111)γ//(110)δ, [111]γ//[110]δ (Kurdjumov–Sachs crystallographic orientation relationship, which is hereafter referred to as the K–S relationship) together with a virtually parallel relationship ((100)γ//(100)δ: preferred growth direction) with a small deviation. Similar relationships are confirmed at other locations of the weld beads. The crystallographic orientation relationship between vermicular δ-ferrite and γ was also examined. The pole figure in Fig. 2 shows a perfect parallel relationship [100]γ//[100]δ with a nearly K–S relationship and a small angle deviation. These observations confirm different orientation relationships between γ and δ-ferrite in the 316FR AF weld metals. These results are included in the previous results for the crystallographic orientation relationship between γ and δ-ferrite in austenitic stainless steel weld metals performed by Inoue et al. [39,40].
Table 1 – Chemical composition range of Type 316FR stainless steel and filler metals (mass %). Materials
C
S
P
Cr
Ni
Mo
Si
Mn
N
Fe
Compositional range of Type 316FR 316FR AF 316FR FA
≦0.020 0.0085 0.0050
≦0.03 0.0009 0.001
0.015–0.045 0.023 0.029
16.00–18.00 17.56 18.51
10.00–14.00 12.02 11.50
2.00–3.00 2.15 2.28
≦1.00 0.44 0.47
≦2.00 0.79 1.50
0.06–0.12 0.088 0.067
Bal. Bal. Bal.
154
M A TE RI A L S CH A RACT ER IZ A TI O N 86 (2 0 1 3 ) 1 5 2–1 6 6
Fig. 1 – EBSD micrographs and pole figures of the cross-section in the 316FR AF weld metal.
4.
Microstructural Changes During Aging
Figs. 3 and 4 show the back-scattered electron (BSE) images from the SEM in the precipitation sequences of the 316FR AF and FA weld metals. In both cases, the precipitates in the initial stage appear in the interior of ferrite, and the number of precipitates increases with aging time. The precipitates are mainly classified as either ones that nucleate at the γ/δ-ferrite interface or ones that precipitate in the interior of the ferrite grains. Fig. 5 presents TEM micrographs of the precipitates. For both weld metals, the dominant precipitates during aging are the sigma (σ: FeCr) and chi (χ: Fe18Cr6Mo5) phases [8]. Furthermore, precipitation of secondary austenite (γ2) is confirmed in both 316FR AF and FA weld metals. Fig. 6 shows a representative γ2 identification, which was observed by EBSD. Consequently, σ, χ, and γ2 mainly precipitate during aging. In
both weld metals, the σ phase preferentially nucleates at the γ/δ-ferrite interface, while the χ phase nucleates in the interior of the ferrite. Further increasing the aging time continuously increases the amount of the σ phase, while the χ phase slightly increases, but then decreases. All the precipitates consist of the σ phase. In other words, once the σ phase precipitates, both precipitates coexist, but the σ phase rapidly grows as the aging time increases. Consequently, the existing χ phase transforms into the σ phase. Figs. 7 and 8 show the EBSD maps and pole figures between δ-ferrite and the χ phase. The crystallographic coherency with the ferrite shows a cube-on-cube orientation relationship ((100)δ//(100)χ, [011]δ//[011]χ). The χ phase in both weld metals preferentially nucleates in the interior of δ-ferrite, while the σ phase nucleates at the γ/δ-ferrite interface, which is typically the preferred precipitation site. As demonstrated by the cube-on-cube relationship, one reason for these behaviors is the crystallographic perfect coherency with δ-ferrite [10–12].
Fig. 2 – EBSD micrographs and pole figures of the cross-section in the 316FR FA weld metal.
M A TE RI A L S C HA RACT ER I ZA TI O N 86 ( 20 1 3 ) 1 5 2–1 6 6
155
Fig. 3 – Back-scattered electron images of SEM showing the microstructural changes during the aging treatment (316FR AF weld metal): (a) aged at 873 K — 10 h, (b) aged at 873 K — 100 h, (c) aged at 973 K — 1 h, (d) aged at 1023 K — 100 h.
5.
Quantitative Evaluation of the Aging Behavior
Figs. 9 and 10 show the area fraction change of each precipitate (σ and χ) for both weld metals. After increasing slightly upon initiation, the amount of the χ phase decreases
as aging time increases for both weld metals. In contrast, the amount of the σ phase continuously increases as aging time increases. The area fraction of the σ phase is saturated at 1.6% (3.9%) in the 316FR AF (316FR FA) weld metal. Because accurately measuring γ2 during aging is extremely difficult, the amount of γ2 is estimated as the difference between the
Fig. 4 – Back-scattered electron images of SEM showing the microstructural changes during the aging treatment (316FR FA weld metal): (a) aged at 873 K — 100 h, (b) aged at 873 K — 500 h, (c) aged at 973 K — 5 h, (d) aged at 973 K — 5 h.
156
M A TE RI A L S CH A RACT ER IZ A TI O N 86 (2 0 1 3 ) 1 5 2–1 6 6
Fig. 5 – TEM micrographs and diffraction patterns of intermetallic phases (σ and χ phases).
Fig. 6 – EBSD micrographs for the analysis of γ2 identification (873 K — 100 h).
saturation points of all the precipitates (2% (5%) in the 316FR AF (316FR FA) weld metal) and the σ phase (1.6% (3.9%) in the 316FR AF (316FR FA) weld metal). Consequently, γ2 at the saturated state is about 0.4% (1.1%) in the 316FR AF (316FR FA) weld metal.
Because the σ and χ phases have been regarded as similar intermetallics with the same negative impact on the mechanical and chemical properties, Fig. 11 compares the precipitation behaviors of the weld metals using the precipitation ratio. Two
Fig. 7 – EBSD micrographs and pole figures for the aged 316FR AF weld metal (1023 K — 1 h).
M A TE RI A L S C HA RACT ER I ZA TI O N 86 ( 20 1 3 ) 1 5 2–1 6 6
157
Fig. 8 – EBSD micrographs and pole figures for the aged 316FR FA weld metal (1023 K — 1 h). main differences are apparent. At aging temperatures of 873 and 923 K, the precipitation initiation time differs between weld metals. Additionally, at other aging temperatures, the
precipitation speed varies. Namely, at the beginning stage of aging at 873 and 923 K, the 316FR AF weld metal has a more rapid onset of precipitation than that of the 316FR FA weld metal.
Fig. 9 – Fractional change of σ, χ phases with aging time at various aging temperatures (316FR AF weld metal).
158
M A TE RI A L S CH A RACT ER IZ A TI O N 86 (2 0 1 3 ) 1 5 2–1 6 6
However, as the aging time and temperature increase, the precipitation speed is reversed, and the speed of the 316FR FA weld metal becomes faster than that of the 316FR AF weld metal.
6.
Kinetics of the Aging Behavior
The kinetics of the precipitation behaviors were examined based on the aforementioned results. A number of previous studies have indicated that the σ phase precipitation behavior in weld metal of austenitic stainless steel can be approximated by that of δ-ferrite decomposition. Herein kinetic approaches were carried out using three different concepts: δ-ferrite decomposition, σ + χ phase precipitation, and σ phase precipitation. Additionally, the possibility of approximating
the σ phase precipitation by δ-ferrite decomposition was examined. Fig. 12 shows the decomposition behaviors of δ-ferrite for both weld metals at different aging temperatures. Regardless of the aging temperature, the decomposed fraction of the ferrite increases sigmoidally in both weld metals as aging time increases. Eventually, the fraction approaches the saturation point. Fig. 11 confirms a similar reversal behavior. Kinetic models and equations have been proposed for various isothermal phase transformation phenomena by employing different impingement exponents, such as the Johnson–Mehl (J–M), the Austin–Rickett (A–R) equation, and the time law for normal grain growth. In particular, the J–M and A–R equations commonly correct certain effects, such as depletion of the solute content in an untransformed matrix due to competitive growth of the reaction products or exhaustion of the nucleation
Fig. 10 – Fractional change of σ, χ phases with the aging time at various aging temperatures (316FR FA weld metal).
M A TE RI A L S C HA RACT ER I ZA TI O N 86 ( 20 1 3 ) 1 5 2–1 6 6
159
Fig. 11 – Comparison of the precipitation (σ + χ) ratio between the 316FR AF and FA weld metals at various aging temperatures.
sites. These equations are generally expressed as an empirical formula for the isothermal transformation of alloys as [41–44] ð1Þ
y ¼ f =f sat n
dy=dt ¼ k t n−1 ð1−yÞm
ð2Þ
k ¼ k0 expð−Q=RT Þ
ð3Þ
where y is the transformed fraction, fsat is the maximum saturation, and f is the amount of precipitation at an arbitrary time. n is the time exponent, which depends on the nucleation mechanism and growth processes, m is an impingement exponent, and t is the holding time. k is a rate constant described by the Arrhenius equation (Eq. (3)), k0 is the pre-exponential factor, Q is the apparent activation energy, and T is the absolute temperature (= aging temperature). To
clarify the correlation between existing equations, an impingement factor (1 − y)m should be introduced. This approach employs an integrable mathematical expression, which indicates that the correlation between transformation behaviors is expressed as і) m = 0; Time law equation n
ð4Þ
y ¼ ðkt Þ
іі) m = 1; Johnson–Mehl equation n y ¼ 1−exp −ðkt Þ
ð5Þ
ііі) m = 2; Austin–Rickett equation n
y=ð1−yÞ ¼ ðkt Þ :
ð6Þ
160
M A TE RI A L S CH A RACT ER IZ A TI O N 86 (2 0 1 3 ) 1 5 2–1 6 6
Fig. 12 – Comparison of the δ-ferrite decomposition ratio between the 316FR AF and FA weld metals at various aging temperatures.
Eqs. (4)–(6) can be respectively transformed with logarithms into Eqs. (7)–(9) as log y ¼ n log t þ log k
ð7Þ
log ln1=1−y ¼ n log t þ log k
ð8Þ
log y=1−y ¼ n log t þ n log k:
ð9Þ
If the plots between log y, log ln 1 / 1 − y, log y / 1 − y, and n log t in Eqs. (7)–(9) can be fitted to a straight line, the time exponent n can be easily evaluated from the slopes of those linear plots. Eq. (3) can be generally transformed into the following Arrhenius form, which determines the activation energy as lnk ¼ ð−Q =RT Þ þ lnk0 :
ð10Þ
Figs. 13–15 respectively show the results of applying the kinetic approaches to δ-ferrite decomposition, σ + χ and σ phase precipitation to both 316FR AF and FA weld metals and their corresponding Arrhenius plots. Regardless of aging temperature, the Johnson–Mehl plot exhibits a good linear relationship, while the other plots fail to adequately describe the relationship. The plots of time law and the Austin–Rickett equation deviate from a linear relationship in the final stages of precipitation, showing a positive and negative curvature, respectively. Tables 3–5 summarize the kinetic parameters for each case. In particular, the kinetics for the ferrite decomposition behavior is similar for both weld metals (similar k0 and Q value between the 316FR AF and FA weld metals in Table 3). However, the kinetics for individual treatments of precipitates (i.e. σ + χ and σ phase precipitation) has different k0 and Q values between weld metals. Namely, the kinetics of
M A TE RI A L S C HA RACT ER I ZA TI O N 86 ( 20 1 3 ) 1 5 2–1 6 6
Fig. 13 – Applicability of kinetic equations for the aging behavior and Arrhenius plot (δ-ferrite decomposition).
Fig. 14 – Applicability of kinetic equations for the aging behavior and Arrhenius plot (σ + χ precipitation).
161
162
M A TE RI A L S CH A RACT ER IZ A TI O N 86 (2 0 1 3 ) 1 5 2–1 6 6
Fig. 15 – Applicability of kinetic equations for the aging behavior and Arrhenius plot (σ precipitation).
σ + χ or σ phase precipitation cannot simply be approximated by the ferrite decomposition behavior. This difference between the kinetics of the ferrite decomposition and individual treatments is attributed to precipitation of γ2, which must be included in ferrite decomposition. For all cases, the activation energies obtained here are similar to the activation energies for the diffusion of the main alloying elements within δ-ferrite (Table 6) [8,16], suggesting that diffusion of the main alloying elements governs the decomposition of δ-ferrite and the precipitation of σ + χ and σ.
initiation time and precipitation growth speed (related to k0 and n in Eq. (4), which are given in Table 4). Namely, the precipitation initiation time of the 316FR AF weld metal is faster than that of the 316FR FA weld metal. Although the precipitation speed of 316FR FA is initially faster than 316FR AF, this behavior is reversed as aging time increases. Typically the difference in precipitation initiation times can be regarded as a variation in the nucleation rates. The concentration of σ and χ phase forming elements (such as Cr and Mo) at or near the nucleation sites governs the nucleation rate [45,46]. Consequently, the distribution state of alloying elements in the ferrite and γ should be significant. In particular, microsegregation of alloying elements occurs during
7. Differences in the σ + χ Phase Precipitation Behaviors Between the 316FR AF and FA Weld Metals The different precipitation behaviors between the 316FR AF and FA weld metals are discussed using the σ + χ phase precipitation behavior (Fig. 11) and its kinetic parameters (Table 4). The precipitation behavior differences between weld metals can be summarized as being due to the precipitation
Table 3 – Determined constants in the Johnson–Mehl equation for δ-ferrite decomposition.
316FR AF weld metal 316FR FA weld metal
n
k0 (/s)
Q (kJ/mol)
0.37 0.77
1.86 × 108 3.11 × 108
237.1 240.1
Table 4 – Determined constants in the Johnson–Mehl equation for σ + χ precipitation. n 316FR AF weld metal 316FR FA weld metal
0.37 0.67
k0 (/s)
Q (kJ/mol) 5
7.71 × 10 6.98 × 106
215.8 226.5
Table 5 – Determined constants in the Johnson–Mehl equation for σ precipitation.
316FR AF weld metal 316FR FA weld metal
n
k0 (/s)
Q (kJ/mol)
0.49 0.82
1.15 × 107 1.40 × 109
242.6 267.8
163
M A TE RI A L S C HA RACT ER I ZA TI O N 86 ( 20 1 3 ) 1 5 2–1 6 6
Table 6 – Activation energies for diffusion of the alloying elements in δ-ferrite. Activation energy (kJ/mol) Diffusion Diffusion Diffusion Diffusion
of of of of
Cr in δ-ferrite Ni in δ-ferrite Mo in δ-ferrite Fe in δ-ferrite
267 262 283 296
solidification of the weld metal for austenitic stainless steel, and the solidification modes in the weld metals possess drastically different microsegregation behaviors [9]. Thus, the influence of microsegregation on the precipitation initiation time cannot be ignored, and so EPMA analysis was employed to assess the distribution of elements (Cr and Mo) for both weld metals. Fig. 16 shows the element profiles of Cr and Mo for the 316FR AF and FA weld metals. Cr and Mo are enriched at the cell or dendrite boundaries where δ-ferrite exists in the 316FR AF weld metal. Although Cr and Mo are also enriched at the cell boundaries (δ-ferrite areas) for the 316FR FA weld metal, the ferrite concentration differs. The concentration distributions were quantitatively compared by point analysis using EPMA. Table 7 shows the measured results for Cr, Mo, Fe, and Ni. All the values in both weld metals agree well with previously reported segregation behaviors for AF and FA mode weld metals [9,47,48]. Simply, segregation of the alloying elements is more severe in the AF mode (316FR AF weld metal) than that in the FA mode (316FR
Table 7 – EPMA analysis results (Point analysis, wt.%). Materials
Phases
Cr
Mo
Ni
Fe
316FR AF weld metal
δ-Ferrite Austenite δ-Ferrite Austenite
25.72 18.17 24.61 19.58
6.78 1.92 4.35 2.41
8.26 12.14 7.63 13.59
61.82 69.37 64.97 66.25
316FR FA weld metal
FA weld metal). Fig. 17 shows the Cr and Mo concentrations in the ferrite and the austenite. One outstanding feature is the concentration difference of Cr and Mo in δ-ferrite; ferrite in the 316FR FA weld metal contains less Cr and Mo than those in the 316FR AF weld metal, despite the fact that the 316FR FA base metal contains more Cr and Mo. Thus, the different precipitation initiation times in the σ + χ phase (Fig. 11) can be explained by the concentrations of Cr and Mo in the ferrite of the weld metal; that is, the rapid precipitation initiation in the 316FR AF weld metal is attributable to the higher concentrations of Cr and Mo in δ-ferrite compared to the 316FR FA weld metal. However, the precipitation speed of the 316FR FA weld metal becomes faster than that of the 316FR AF weld metal, and eventually the rate is reversed. This behavior is attributed to і) the different concentrations of Cr and Mo in the γ vicinity of the ferrite and іі) variation in the nucleation sites of the precipitates (amount of ferrite). In the as-welded situation, the Cr and Mo contents in δ-ferrite of the 316FR FA weld metal are smaller than those in the 316FR AF weld metal. During the
Fig. 16 – Microsegregation behavior of the alloying elements using EPMA for the 316FR AF and FA weld metals.
164
M A TE RI A L S CH A RACT ER IZ A TI O N 86 (2 0 1 3 ) 1 5 2–1 6 6
Fig. 17 – Point analysis of the microsegregation behavior using EPMA for the 316FR AF and FA weld metals.
aging treatment, Cr and Mo must diffuse from γ into δ-ferrite because the diffusion rates of Cr and Mo in ferrite are about 100 times faster than those in γ [49]. Fig. 17 also shows Cr and Mo concentrations in γ for both weld metals. γ for the 316FR FA weld metal has a higher concentration than that in the 316FR AF weld metal. Despite the initially slower precipitation initiation of the 316FR FA weld metal, diffusion of Cr and Mo from γ into δ-ferrite must occur as precipitation progresses because the 316FR FA weld metal, which possesses more Cr and Mo in γ provides more opportunities for precipitation events. Moreover, the available nucleation sites of the σ and χ phases (e.g., the ferrite area of the 316FR FA weld metal) are greater than those in the AF weld metal. Consequently, the rapid precipitation speed in the 316FR FA weld metal is due to і) large concentrations of Cr and Mo in γ and іі) the large amount of ferrite. Furthermore, the kinetic parameters (larger values for k0 and n in the 316FR FA than those of the 316FR AF weld metal) (Table 4) are
consistent with the rapid precipitation speed in the 316FR FA weld metal.
8. Prediction of the Aging Behavior Between 773 and 823 K for FBR Service Based on the determined kinetic equations for both weld metals, the decomposition behavior of δ-ferrite and the precipitation (σ + χ phases) behavior were predicted under the service conditions of a FBR (773 and 823 K) (Fig. 18). Initially, the 316FR FA weld metal has a lower decomposition rate of δ-ferrite than that of the 316FR AF weld metal, but the rate of the 316FR FA weld metal eventually exceeds that of the 316FR AF weld metal as operations progress. Table 8 lists the decomposition values of both weld materials at select aging times. Both weld metals have decomposition rates for δ-ferrite at an 823 K service temperature that are about 15 times faster than that at 773 K.
Fig. 18 – Prediction results: δ-ferrite decomposition ratio and σ + χ phase precipitation at the service temperatures of a FBR (773 and 823 K).
165
M A TE RI A L S C HA RACT ER I ZA TI O N 86 ( 20 1 3 ) 1 5 2–1 6 6
Fig. 18 and Table 9 show the prediction results for the σ + χ precipitation. Initially, the amount of σ + χ phases is similar, regardless of service temperature and weld metal. However, after 0.2 years of service, the amount of the σ + χ phases in the 316FR FA weld metal suddenly increases compared to that in the 316FR AF weld metal.
9.
Conclusions
Herein the microstructural changes and kinetic approaches of precipitates (σ and χ) during aging treatment for two types of weld metals of 316FR stainless steel (316FR AF and FA) were examined. The main conclusions are summarized below: (1) The 316FR weld metals contain approx. 2% (5%) of δ-ferrite formed during the AF (FA) solidification mode. The main precipitates in δ-ferrite are σ, χ, and γ2, which nucleate at the inside or surrounding δ-ferrite during aging treatments at 823 to 1073 K. These precipitates consume all the ferrite for both weld metals. (2) The decomposition behavior of δ-ferrite (δ-ferrite decomposition, σ + χ precipitation, σ precipitation) in the 316FR AF and FA weld metals was examined using three different kinetic approaches: the time law, Johnson– Mehl, and Austin–Rickett equations. The Johnson– Mehl-type equation is the most applicable to the decomposition behavior of the ferrite. The precipitation behaviors of σ and χ phases can be controlled by the diffusion of main alloying elements (Cr, Ni, and Mo) in δ-ferrite. (3) Precipitation (σ + χ) between both solidification modes differs due to the precipitation initiation time and precipitation speed. The different concentrations of Cr and Mo in δ-ferrite in the as-welded situation explain the difference in the precipitation initiation time, while the difference in the precipitation speed is presumed to be due to the varying concentrations of Cr and Mo in γ and the different amount of nucleation sites (i.e., amount of δ-ferrite). (4) Decomposition of δ-ferrite and σ + χ precipitation in the 316FR AF and FA weld metals were predicted at practical operating temperatures of a FBR (773 and 823 K) using the determined Johnson–Mehl-type equation. The predicted decomposition rate of δ-ferrite
Table 8 – Prediction results: decomposed fraction of δ-ferrite for select times in Fig. 18. Aging time (year)
0.1 1 5 10 50 100
Decomposed fraction of δ-ferrite 773 K
823 K
316FR AF
316FR FA
316FR AF
316FR FA
0.28 0.55 0.76 0.84 0.96 0.98
0.10 0.47 0.89 0.97 1.00 1.00
0.54 0.84 0.96 0.98 0.99 1.00
0.47 0.97 1.00 1.00 1.00 1.00
Table 9 – Prediction results: area fraction of σ + χ for select times in Fig. 18. Area fraction of σ + χ (%)
Aging time (year)
773 K
823 K
316FR AF
316FR FA
316FR AF
316FR FA
0.22 0.47 0.76 0.90 1.26 1.38
0.17 0.79 1.89 2.55 3.75 3.87
0.44 0.85 1.19 1.33 1.54 1.59
0.71 2.39 3.67 3.85 3.90 3.90
0.1 1 5 10 50 100
at an 823 K service temperature is about 15 times faster than that at 773 K for both weld metals. Furthermore, in the initial service stage, the amount of σ + χ phases is similar, regardless of service temperature for both weld metals, but after 0.2 years of operations, the amount of σ + χ phases in the 316FR FA weld metal drastically increases compared to the 316FR AF weld metal.
Acknowledgment The present study includes the result of the “Core R&D program for commercialization of the fast breeder reactor by utilizing Monju” entrusted to the University of Fukui by the Ministry of Education, Culture, Sports, Science and Technology of Japan (MEXT).
REFERENCES [1] Iida K, Asada Y, Okabayashi K, Nagata T. Construction codes developed for prototype FBR Monju. Nucl Eng Des 1987;98:283–8. [2] Iida K, Asada Y, Okabayashi K, Nagata T. Simplified analysis and design for elevated temperature components of Monju. Nucl Eng Des 1987;98:305–17. [3] Nakazawa T, Kimura H, Kimura K, Kaguchi H. Advanced type stainless steel 316FR for fast breeder reactor structures. J Mater Process Technol 2003;143:905–9. [4] Takahashi Y, Shibamoto H, Inoue K. Study on creep-fatigue life prediction methods for low-carbon nitrogen-controlled 316 stainless steel (316FR). Nucl Eng Des 2008;238:322–35. [5] Furukawa T, Kato S, Yoshida E. Compatibility of FBR materials with sodium. J Nucl Mater 2009;392:249–54. [6] Nogami S, Hasegawa A, Tanno T, Imasaki K, Abe K. High-temperature helium embrittlement of 316FR steel. J Nucl Sci Technol 2011;48:130–4. [7] Konomura M, Ichimiya M. Re-weldability tests of irradiated 316L(N) stainless steel using laser welding technique. J Nucl Mater 2002;307-11:1584–9. [8] Chun EJ, Baba H, Terashima K, Nishimoto K, Saida K. Prediction of σ phase embrittlement in type 316FR weld metal. Weld J May 2013:133–9. [9] Kou S. Welding metallurgy. 2nd ed. John Wiley & Sons Inc.; 2003 216–300. [10] Bae JI, Kim ST, Lee TH, Ha HY, Kim SJ, Park YH. High temperature precipitation behavior of high-nitrogen duplex stainless steel. Korean J Met Mater 2010;49:99–103.
166
M A TE RI A L S CH A RACT ER IZ A TI O N 86 (2 0 1 3 ) 1 5 2–1 6 6
[11] Byun SH, Kang N, Lee TH, Ahn SK, Lee HW, Chang WS , et al. Kinetics of Cr/Mo-rich precipitates formation for 25Cr–6.9Ni–3.8Mo–0.3N super duplex stainless steel. Met Mater Int 2012;1:201–7. [12] Lee TH, Ha HY, Hwang B, Kim SJ. Isothermal decomposition of ferrite in a high-nitrogen, nickel-free duplex stainless steel. Metall Mater Trans A 2012;43A:822–32. [13] Nakao Y, Nishimoto K, Inoue S. Kinetics of the formation of austenite at constant temperature — study on the phase transformation in welded joints of duplex stainless steels (report 1). J Jpn Weld Soc 1981;50:514–20. [14] Wasnik DN, Dey GK, Kain V, Samajdar I. Precipitation stages in a 316L austenitic stainless steel. Scr Mater 2003;49:135–41. [15] Barcik J. The kinetics of σ-phase precipitation in AISI 310 and AISI 316 steels. Metall Trans A 1983;14:635–41. [16] Magnabosco R. Kinetics of sigma phase formation in a duplex stainless steel. Mater Res 2009;12:321–7. [17] Nilsson JO, Wilson A. Influence of isothermal phase transformations on toughness and pitting corrosion of super duplex stainless steel SAF 2507. Mater Sci Technol 1993;9:545–54. [18] Nilsson JO, Huhtala T, Jonsson P, Karlsson L, Wilson A. Structural stability of super duplex stainless steel weld metals and its dependence on tungsten and copper. Metall Mater Trans A 1996;27A:2196–208. [19] Lee YH, Kim KT, Lee YD, Kim KY. Effects of W substitution on σ and χ phase precipitation and toughness in duplex stainless steels. Mater Sci Technol 1998;14:757–64. [20] Kim SB, Paik KW, Kim YG. Effect of Mo substitution by W on high temperature embrittlement characteristics in duplex stainless steels. Mater Sci Eng A 1998;247:67–74. [21] Chen TH, Weng KL, Yang JR. The effect of high temperature exposure on the microstructural stability and toughness property in a 2205 duplex stainless steel. Mater Sci Eng A 2002;338:259–70. [22] Michalska J, Sozanska M. Qualitative and quantitative analysis of σ and χ phases in 2205 duplex stainless steel. Mater Charact 2006;56:355–62. [23] Kim YJ, Ugurlu O, Jiang C, Gleeson B. Microstructural evolution of secondary phase in the cast duplex stainless steels CD3MN and CD3MWCuN. Metall Mater Trans A 2007;38A:203–11. [24] Escriba DM, Materna-Morris E, Plaut RL, Padilha AF. Chi-phase precipitation in a duplex stainless steel. Mater Charact 2009;60:1214–9. [25] Huang CS, Shih CC. Effect of nitrogen and high temperature aging on σ phase precipitation of duplex stainless steel. Mater Sci Eng A 2005;402:66–75. [26] Pardal JM, Tavares SSM, Fonseca MDPC, Souza JAS, Vieira LM, Abreu HFGD. Deleterious phase precipitation on superduplex stainless steel UNS S32750 characterization by light optical and scanning electron microscopy. Mater Res 2010;13:401–7. [27] Calliari I, Straffelini G, Ramous E. Investigation of secondary phase effect on 2205 DSS fracture toughness. Mater Sci Technol 2010;26:81–6. [28] Calliari I, Pellizzari M, Zanellato M, Ramous E. The phase stability in Cr–Ni and Cr–Mn duplex stainless steels. J Mater Sci 2011;46:6916–24. [29] Kashiwar A, Vennela NP, Kamath SL, Khatirkar RK. Effect of solution annealing temperature on precipitation in 2205 duplex stainless steel. Mater Charact 2012;74:55–63. [30] Pohl M, Storz O, Glogowski T. Effect of intermetallic precipitation on the properties of duplex stainless steel. Mater Charact 2007;58:65–71.
[31] Ghosh SK, Mondal S. High temperature aging behaviour of a duplex stainless steel. Mater Charact 2008;59:1776–83. [32] Gill TPS, Vijayalakshmi M, Rodriguez P, Padmanabhan KA. On microstructure–property correlation of thermally aged type 316L stainless steel weld metal. Metall Trans A 1989;20:1115–24. [33] Gill TPS, Shankar V, Vijayalakshmi M, Rodriguez P. Influence of carbon on the transformation kinetics of delta-ferrite in type 316 stainless steel weld metals. Scr Metall Mater 1992;27:313–8. [34] Gill TPS, Shankar V, Pujar MG, Rodriguez P. Effect of composition on the transformation of δ to σ in type 316 stainless steel weld metals. Scr Metall Mater 1995;10:1595–600. [35] Mathew MD, Sasikala G, Mannan SL, Rodriguez P. Effect of stress on the transformation behaviour of delta-ferrite in type 316 stainless steel weld metal. J Mater Sci Lett 1994;13:194–6. [36] Sasikala G, Ray SK, Mannan SL. Kinetics of transformation of delta ferrite during creep in a type 316(N) stainless steel weld metal. Mater Sci Eng A 2003;359:86–90. [37] Chun EJ, Baba H, Terashima K, Nishimoto K, Saida K. Prediction of σ phase precipitation in type 316FR stainless steel weld metal. J Jpn Weld Soc 2013;31:168s–72s. [38] Badji R, Bouabdallah M, Bacroix B, Kahloun C, Berrahar K, Kherrouba N. Effect of solution temperature on the precipitation kinetic of σ-phase in 2205 duplex stainless steel welds. Mater Sci Eng A 2008;469:447–54. [39] Inoue H, Koseki T, Okita S, Fuji M. Solidification and transformation behavior of austenitic stainless steel weld metals solidified as primary austenite — study of solidification and subsequent transformation of Cr–Ni stainless steel weld metals (1st report). Weld Int 1997;11:876–87. [40] Inoue H, Koseki T, Okita S, Fuji M. Solidification and transformation behavior of austenitic stainless steel weld metals solidified as primary ferrite — study of solidification and subsequent transformation of Cr–Ni stainless steel weld metals (2nd report). Weld Int 1997;11:937–49. [41] Johnson WA, Mehl RF. Reaction kinetics in processes of nucleation and growth. Trans AIME 1939;135:416–59. [42] Starink MJ. Kinetic equations for diffusion-controlled precipitation reactions. J Mater Sci 1997;32:4061–70. [43] Nakao Y, Nishimoto K, Ohtani K. Transformation kinetics of δ ferrite during heat treatment. J Jpn Weld Soc 1984;2:61–7. [44] Lee ES, Kim YG. A transformation kinetic model and its application to Cu–Zn–Al shape memory alloys—1. Isothermal conditions. Acta Metall Mater 1990;38:1669–76. [45] Porter DA, Easterling KE, Sherif MY. Phase transformations in metals and alloys. 3rd ed. New York: CRC Press; 2009. [46] Reed-Hill RE, Ababschian R. Physical metallurgy principles. 3rd ed. Boston: PWS Publishing Company; 1992. [47] Brooks JA, Williams JC, Thompson AW. STEM analysis of primary austenite solidified stainless steel welds. Metall Trans A 1983;14A:1271–81. [48] Brooks JA, Williams JC, Thompson AW. Microstructural origin of the skeletal ferrite morphology of austenitic stainless steel welds. Metall Trans A 1983;14A:23–31. [49] Hsieh CC, Wu W. Discussing the precipitation behavior of σ phase using diffusion equation and thermodynamic simulation in dissimilar stainless steel. J Alloys Compd 2010;506:820–5.
Available online at www.sciencedirect.com
ScienceDirect www.elsevier.com/locate/matchar
Precipitation of sigma and chi phases in δ-ferrite of Type 316FR weld metals Eun Joon Chuna,⁎, Hayato Babaa , Kazutoshi Nishimotob , Kazuyoshi Saidaa a Division of Materials and Manufacturing Science, Graduate School of Engineering, Osaka University, 2-1 Yamada-oka, Suita-shi, Osaka 565-0871, Japan b Department of the Application of Nuclear Technology, Fukui University of Technology, Gakuen 3-6-1, Fukui-shi, Fukui 910-8505, Japan
AR TIC LE D ATA
ABSTR ACT
Article history:
The decomposition behavior and kinetics of δ-ferrite are examined using aging treatments
Received 23 July 2013
between 873 and 1073 K for Type 316FR stainless steel weld metals with different
Received in revised form 25
solidification modes (316FR AF, 316FR FA). The dominant precipitates are sigma, chi, and
September 2013
secondary austenite nucleated at δ-ferrite/austenite interfaces or in the interior of the
Accepted 2 October 2013
ferrite grains. These precipitates consume all the ferrite during isothermal aging in both 316FR AF and FA weld metals. Differences in the precipitation behavior (precipitation initiation time and precipitation speed) between weld metals can be explained by і) the
Keywords:
degree of Cr and Mo microsegregation within δ-ferrite or austenite near ferrite and іі) the
Type 316FR
nucleation sites induced due to the solidification mode (AF or FA), such as the ferrite
Weld metal
amount. For both weld materials, a Johnson–Mehl-type equation can express the pre-
Delta ferrite decomposition
cipitation behavior of the sigma + chi phases and quantitatively predict the behavior at the
Sigma phase precipitation
service-exposure temperatures of a fast breed reactor.
Kinetics
© 2013 Elsevier Inc. All rights reserved.
Prediction
1.
Introduction
Compared to common austenitic stainless steels, an advanced Type 316FR austenitic stainless steel has an improved creep rupture resistance and is a key component in the nextgeneration of commercialized fast breed reactors (FBRs) [1–6]. Because a FBR typically operates for about 60 years at high temperatures (773–823 K) due to its outstanding ability to generate heat compared to other types of nuclear reactors (ex. light water-cooled reactor), aging of the structural materials is an inevitable issue. Thus, before FBRs can be commercialized, the aging behavior and the appropriate repair processes must be clarified [7]. One of the most important factors to determine the aged degree of components made of austenitic stainless steels is the inferior microstructural stability of the weld metal at ele-
⁎ Corresponding author. Tel./fax: + 81 6 6879 7543. E-mail address: [email protected] (E.J. Chun). 1044-5803/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.matchar.2013.10.003
vated temperatures. Weld metals of austenitic stainless steel generally involve some amount of δ-ferrite, which is used to prevent solidification cracking during welding. Weld metals are typically composed of different solidification modes, such as primary austenite (γ) with a secondary phase δ-ferrite solidification mode (AF mode) or primary δ-ferrite with a secondary phase γ solidification mode (FA mode) according to the Creq/Nieq variation. However, δ-ferrite may decompose into intermetallic phases (sigma or chi) in a high temperature environment, negatively impacting the mechanical and chemical properties of a weld joint [8]. Hot cracking is a common problem while welding austenitic stainless steel in the AF mode [9]. Because changing the solidification mode or optimizing the δ-ferrite content during solidification can control the hot cracking susceptibility [9], the
153
M A TE RI A L S C HA RACT ER I ZA TI O N 86 ( 20 1 3 ) 1 5 2–1 6 6
chemical composition of the consumable welding components can be adjusted in the FA mode. However excessively increasing the amount of ferrite intensifies the embrittlement of the weld metal because a large amount of intermetallics is transformed while in actual service. Although numerous studies have investigated the aging behavior (precipitation) in austenitic and duplex stainless steels [10–38], information about the kinetics for the weld metal is insufficient compared to that for the base metal [32–38]. Most studies have approximated the kinetics for δ-ferrite decomposition without employing a detailed approach to evaluate individual precipitates [32–38]. Moreover, comparison studies on the precipitation kinetics in weld metals between the AF and FA modes with similar composition ranges are insufficient, despite the possibility that the kinetics may depend on microsegregation according to solidification mode variations. The present study aims to clarify the δ-ferrite decomposition behaviors in Type 316FR stainless steel weld metals utilizing welding consumables with two different solidification modes (AF and FA modes). The decomposition behaviors of ferrite are quantified by accurately measuring the amount of each precipitate. Based upon quantitative analyses, kinetic approaches are pursued and different precipitation behaviors between welding consumables are discussed.
2.
Materials and Experimental Procedure
The materials used are within the compositional range of Type 316FR stainless steel. Table 1 lists their chemical compositions. The designed steel with the AF mode is referred to as 316FR AF, while that with the FA mode is referred to as 316FR FA. The base metal plates of these steels, which measured 120 mm × 40 mm × 3 mm, were manufactured via hot-rolling. To prepare the weld metal, bead-on-plate welding was performed using gas tungsten arc welding (GTAW) on the manufactured plates with an arc current of 110 A, arc voltage of 14 V, and a welding speed of 1.67 mm/s. Welded specimens were aged at various conditions (Table 2) and subsequently quenched with water. The as-welded microstructures were observed by electron backscatter diffraction (EBSD) after electrolytic etching with a 10% aqueous solution of KOH. To investigate the elemental distribution in the as-welded state, an electron probe X-ray micro analyzer (EMPA) measured the concentration profiles of the elements. Precipitates were identified by transmission electron microscopy (TEM) under an acceleration voltage of 200 kV after jet polishing with a solution of perchloric acid (5%) and acetic acid (95%) using an applied voltage of 50 V. To clarify the microstructural changes due to the aging treatment, specimens were observed by scanning electron
Table 2 – Aging conditions. Temperature (K) 873 923 973 1023 1073
Aging time (h) 0.5 0.5 0.5 0.5 0.5
1 1 1 1 1
5 5 5 5 5
10 10 10 10 10
50 50 50 50 50
100 100 100 100 100
500 500 500 500 –
1000 1000 1000 1000 –
microscopy (SEM) and EBSD using both electrolytic etching and mechanical polishing with colloidal silica. Image processing at a 1000-time magnification on the basis of the SEM micrographs quantitatively measured the area fraction of the precipitates.
3. Characterization of Weld Microstructures and Solidification Modes Fig. 1, which shows the as-welded microstructure of the 316FR AF weld metals, confirms the cellular morphology associated with the solidification behavior. All the δ-ferrite is located at the cell boundaries or triple points of γ with an elongated or globular shape. The average volume fraction of ferrite is 2%. These characteristics are consistent with the AF mode [39]. Fig. 2 shows the microstructure of 316FR FA after GTAW bead-on-plate. The average volume fraction of δ-ferrite is 5%. The shape associated with the bands of vermicular δ-ferrite enclosed by γ at the center of a dendritic cell is consistent with the FA mode [40]. Fig. 1 also shows the corresponding pole figures to examine the crystallographic orientation relationship between γ and δ-ferrite. The preferred orientation relationship between γ and δ-ferrite is (111)γ//(110)δ, [111]γ//[110]δ (Kurdjumov–Sachs crystallographic orientation relationship, which is hereafter referred to as the K–S relationship) together with a virtually parallel relationship ((100)γ//(100)δ: preferred growth direction) with a small deviation. Similar relationships are confirmed at other locations of the weld beads. The crystallographic orientation relationship between vermicular δ-ferrite and γ was also examined. The pole figure in Fig. 2 shows a perfect parallel relationship [100]γ//[100]δ with a nearly K–S relationship and a small angle deviation. These observations confirm different orientation relationships between γ and δ-ferrite in the 316FR AF weld metals. These results are included in the previous results for the crystallographic orientation relationship between γ and δ-ferrite in austenitic stainless steel weld metals performed by Inoue et al. [39,40].
Table 1 – Chemical composition range of Type 316FR stainless steel and filler metals (mass %). Materials
C
S
P
Cr
Ni
Mo
Si
Mn
N
Fe
Compositional range of Type 316FR 316FR AF 316FR FA
≦0.020 0.0085 0.0050
≦0.03 0.0009 0.001
0.015–0.045 0.023 0.029
16.00–18.00 17.56 18.51
10.00–14.00 12.02 11.50
2.00–3.00 2.15 2.28
≦1.00 0.44 0.47
≦2.00 0.79 1.50
0.06–0.12 0.088 0.067
Bal. Bal. Bal.
154
M A TE RI A L S CH A RACT ER IZ A TI O N 86 (2 0 1 3 ) 1 5 2–1 6 6
Fig. 1 – EBSD micrographs and pole figures of the cross-section in the 316FR AF weld metal.
4.
Microstructural Changes During Aging
Figs. 3 and 4 show the back-scattered electron (BSE) images from the SEM in the precipitation sequences of the 316FR AF and FA weld metals. In both cases, the precipitates in the initial stage appear in the interior of ferrite, and the number of precipitates increases with aging time. The precipitates are mainly classified as either ones that nucleate at the γ/δ-ferrite interface or ones that precipitate in the interior of the ferrite grains. Fig. 5 presents TEM micrographs of the precipitates. For both weld metals, the dominant precipitates during aging are the sigma (σ: FeCr) and chi (χ: Fe18Cr6Mo5) phases [8]. Furthermore, precipitation of secondary austenite (γ2) is confirmed in both 316FR AF and FA weld metals. Fig. 6 shows a representative γ2 identification, which was observed by EBSD. Consequently, σ, χ, and γ2 mainly precipitate during aging. In
both weld metals, the σ phase preferentially nucleates at the γ/δ-ferrite interface, while the χ phase nucleates in the interior of the ferrite. Further increasing the aging time continuously increases the amount of the σ phase, while the χ phase slightly increases, but then decreases. All the precipitates consist of the σ phase. In other words, once the σ phase precipitates, both precipitates coexist, but the σ phase rapidly grows as the aging time increases. Consequently, the existing χ phase transforms into the σ phase. Figs. 7 and 8 show the EBSD maps and pole figures between δ-ferrite and the χ phase. The crystallographic coherency with the ferrite shows a cube-on-cube orientation relationship ((100)δ//(100)χ, [011]δ//[011]χ). The χ phase in both weld metals preferentially nucleates in the interior of δ-ferrite, while the σ phase nucleates at the γ/δ-ferrite interface, which is typically the preferred precipitation site. As demonstrated by the cube-on-cube relationship, one reason for these behaviors is the crystallographic perfect coherency with δ-ferrite [10–12].
Fig. 2 – EBSD micrographs and pole figures of the cross-section in the 316FR FA weld metal.
M A TE RI A L S C HA RACT ER I ZA TI O N 86 ( 20 1 3 ) 1 5 2–1 6 6
155
Fig. 3 – Back-scattered electron images of SEM showing the microstructural changes during the aging treatment (316FR AF weld metal): (a) aged at 873 K — 10 h, (b) aged at 873 K — 100 h, (c) aged at 973 K — 1 h, (d) aged at 1023 K — 100 h.
5.
Quantitative Evaluation of the Aging Behavior
Figs. 9 and 10 show the area fraction change of each precipitate (σ and χ) for both weld metals. After increasing slightly upon initiation, the amount of the χ phase decreases
as aging time increases for both weld metals. In contrast, the amount of the σ phase continuously increases as aging time increases. The area fraction of the σ phase is saturated at 1.6% (3.9%) in the 316FR AF (316FR FA) weld metal. Because accurately measuring γ2 during aging is extremely difficult, the amount of γ2 is estimated as the difference between the
Fig. 4 – Back-scattered electron images of SEM showing the microstructural changes during the aging treatment (316FR FA weld metal): (a) aged at 873 K — 100 h, (b) aged at 873 K — 500 h, (c) aged at 973 K — 5 h, (d) aged at 973 K — 5 h.
156
M A TE RI A L S CH A RACT ER IZ A TI O N 86 (2 0 1 3 ) 1 5 2–1 6 6
Fig. 5 – TEM micrographs and diffraction patterns of intermetallic phases (σ and χ phases).
Fig. 6 – EBSD micrographs for the analysis of γ2 identification (873 K — 100 h).
saturation points of all the precipitates (2% (5%) in the 316FR AF (316FR FA) weld metal) and the σ phase (1.6% (3.9%) in the 316FR AF (316FR FA) weld metal). Consequently, γ2 at the saturated state is about 0.4% (1.1%) in the 316FR AF (316FR FA) weld metal.
Because the σ and χ phases have been regarded as similar intermetallics with the same negative impact on the mechanical and chemical properties, Fig. 11 compares the precipitation behaviors of the weld metals using the precipitation ratio. Two
Fig. 7 – EBSD micrographs and pole figures for the aged 316FR AF weld metal (1023 K — 1 h).
M A TE RI A L S C HA RACT ER I ZA TI O N 86 ( 20 1 3 ) 1 5 2–1 6 6
157
Fig. 8 – EBSD micrographs and pole figures for the aged 316FR FA weld metal (1023 K — 1 h). main differences are apparent. At aging temperatures of 873 and 923 K, the precipitation initiation time differs between weld metals. Additionally, at other aging temperatures, the
precipitation speed varies. Namely, at the beginning stage of aging at 873 and 923 K, the 316FR AF weld metal has a more rapid onset of precipitation than that of the 316FR FA weld metal.
Fig. 9 – Fractional change of σ, χ phases with aging time at various aging temperatures (316FR AF weld metal).
158
M A TE RI A L S CH A RACT ER IZ A TI O N 86 (2 0 1 3 ) 1 5 2–1 6 6
However, as the aging time and temperature increase, the precipitation speed is reversed, and the speed of the 316FR FA weld metal becomes faster than that of the 316FR AF weld metal.
6.
Kinetics of the Aging Behavior
The kinetics of the precipitation behaviors were examined based on the aforementioned results. A number of previous studies have indicated that the σ phase precipitation behavior in weld metal of austenitic stainless steel can be approximated by that of δ-ferrite decomposition. Herein kinetic approaches were carried out using three different concepts: δ-ferrite decomposition, σ + χ phase precipitation, and σ phase precipitation. Additionally, the possibility of approximating
the σ phase precipitation by δ-ferrite decomposition was examined. Fig. 12 shows the decomposition behaviors of δ-ferrite for both weld metals at different aging temperatures. Regardless of the aging temperature, the decomposed fraction of the ferrite increases sigmoidally in both weld metals as aging time increases. Eventually, the fraction approaches the saturation point. Fig. 11 confirms a similar reversal behavior. Kinetic models and equations have been proposed for various isothermal phase transformation phenomena by employing different impingement exponents, such as the Johnson–Mehl (J–M), the Austin–Rickett (A–R) equation, and the time law for normal grain growth. In particular, the J–M and A–R equations commonly correct certain effects, such as depletion of the solute content in an untransformed matrix due to competitive growth of the reaction products or exhaustion of the nucleation
Fig. 10 – Fractional change of σ, χ phases with the aging time at various aging temperatures (316FR FA weld metal).
M A TE RI A L S C HA RACT ER I ZA TI O N 86 ( 20 1 3 ) 1 5 2–1 6 6
159
Fig. 11 – Comparison of the precipitation (σ + χ) ratio between the 316FR AF and FA weld metals at various aging temperatures.
sites. These equations are generally expressed as an empirical formula for the isothermal transformation of alloys as [41–44] ð1Þ
y ¼ f =f sat n
dy=dt ¼ k t n−1 ð1−yÞm
ð2Þ
k ¼ k0 expð−Q=RT Þ
ð3Þ
where y is the transformed fraction, fsat is the maximum saturation, and f is the amount of precipitation at an arbitrary time. n is the time exponent, which depends on the nucleation mechanism and growth processes, m is an impingement exponent, and t is the holding time. k is a rate constant described by the Arrhenius equation (Eq. (3)), k0 is the pre-exponential factor, Q is the apparent activation energy, and T is the absolute temperature (= aging temperature). To
clarify the correlation between existing equations, an impingement factor (1 − y)m should be introduced. This approach employs an integrable mathematical expression, which indicates that the correlation between transformation behaviors is expressed as і) m = 0; Time law equation n
ð4Þ
y ¼ ðkt Þ
іі) m = 1; Johnson–Mehl equation n y ¼ 1−exp −ðkt Þ
ð5Þ
ііі) m = 2; Austin–Rickett equation n
y=ð1−yÞ ¼ ðkt Þ :
ð6Þ
160
M A TE RI A L S CH A RACT ER IZ A TI O N 86 (2 0 1 3 ) 1 5 2–1 6 6
Fig. 12 – Comparison of the δ-ferrite decomposition ratio between the 316FR AF and FA weld metals at various aging temperatures.
Eqs. (4)–(6) can be respectively transformed with logarithms into Eqs. (7)–(9) as log y ¼ n log t þ log k
ð7Þ
log ln1=1−y ¼ n log t þ log k
ð8Þ
log y=1−y ¼ n log t þ n log k:
ð9Þ
If the plots between log y, log ln 1 / 1 − y, log y / 1 − y, and n log t in Eqs. (7)–(9) can be fitted to a straight line, the time exponent n can be easily evaluated from the slopes of those linear plots. Eq. (3) can be generally transformed into the following Arrhenius form, which determines the activation energy as lnk ¼ ð−Q =RT Þ þ lnk0 :
ð10Þ
Figs. 13–15 respectively show the results of applying the kinetic approaches to δ-ferrite decomposition, σ + χ and σ phase precipitation to both 316FR AF and FA weld metals and their corresponding Arrhenius plots. Regardless of aging temperature, the Johnson–Mehl plot exhibits a good linear relationship, while the other plots fail to adequately describe the relationship. The plots of time law and the Austin–Rickett equation deviate from a linear relationship in the final stages of precipitation, showing a positive and negative curvature, respectively. Tables 3–5 summarize the kinetic parameters for each case. In particular, the kinetics for the ferrite decomposition behavior is similar for both weld metals (similar k0 and Q value between the 316FR AF and FA weld metals in Table 3). However, the kinetics for individual treatments of precipitates (i.e. σ + χ and σ phase precipitation) has different k0 and Q values between weld metals. Namely, the kinetics of
M A TE RI A L S C HA RACT ER I ZA TI O N 86 ( 20 1 3 ) 1 5 2–1 6 6
Fig. 13 – Applicability of kinetic equations for the aging behavior and Arrhenius plot (δ-ferrite decomposition).
Fig. 14 – Applicability of kinetic equations for the aging behavior and Arrhenius plot (σ + χ precipitation).
161
162
M A TE RI A L S CH A RACT ER IZ A TI O N 86 (2 0 1 3 ) 1 5 2–1 6 6
Fig. 15 – Applicability of kinetic equations for the aging behavior and Arrhenius plot (σ precipitation).
σ + χ or σ phase precipitation cannot simply be approximated by the ferrite decomposition behavior. This difference between the kinetics of the ferrite decomposition and individual treatments is attributed to precipitation of γ2, which must be included in ferrite decomposition. For all cases, the activation energies obtained here are similar to the activation energies for the diffusion of the main alloying elements within δ-ferrite (Table 6) [8,16], suggesting that diffusion of the main alloying elements governs the decomposition of δ-ferrite and the precipitation of σ + χ and σ.
initiation time and precipitation growth speed (related to k0 and n in Eq. (4), which are given in Table 4). Namely, the precipitation initiation time of the 316FR AF weld metal is faster than that of the 316FR FA weld metal. Although the precipitation speed of 316FR FA is initially faster than 316FR AF, this behavior is reversed as aging time increases. Typically the difference in precipitation initiation times can be regarded as a variation in the nucleation rates. The concentration of σ and χ phase forming elements (such as Cr and Mo) at or near the nucleation sites governs the nucleation rate [45,46]. Consequently, the distribution state of alloying elements in the ferrite and γ should be significant. In particular, microsegregation of alloying elements occurs during
7. Differences in the σ + χ Phase Precipitation Behaviors Between the 316FR AF and FA Weld Metals The different precipitation behaviors between the 316FR AF and FA weld metals are discussed using the σ + χ phase precipitation behavior (Fig. 11) and its kinetic parameters (Table 4). The precipitation behavior differences between weld metals can be summarized as being due to the precipitation
Table 3 – Determined constants in the Johnson–Mehl equation for δ-ferrite decomposition.
316FR AF weld metal 316FR FA weld metal
n
k0 (/s)
Q (kJ/mol)
0.37 0.77
1.86 × 108 3.11 × 108
237.1 240.1
Table 4 – Determined constants in the Johnson–Mehl equation for σ + χ precipitation. n 316FR AF weld metal 316FR FA weld metal
0.37 0.67
k0 (/s)
Q (kJ/mol) 5
7.71 × 10 6.98 × 106
215.8 226.5
Table 5 – Determined constants in the Johnson–Mehl equation for σ precipitation.
316FR AF weld metal 316FR FA weld metal
n
k0 (/s)
Q (kJ/mol)
0.49 0.82
1.15 × 107 1.40 × 109
242.6 267.8
163
M A TE RI A L S C HA RACT ER I ZA TI O N 86 ( 20 1 3 ) 1 5 2–1 6 6
Table 6 – Activation energies for diffusion of the alloying elements in δ-ferrite. Activation energy (kJ/mol) Diffusion Diffusion Diffusion Diffusion
of of of of
Cr in δ-ferrite Ni in δ-ferrite Mo in δ-ferrite Fe in δ-ferrite
267 262 283 296
solidification of the weld metal for austenitic stainless steel, and the solidification modes in the weld metals possess drastically different microsegregation behaviors [9]. Thus, the influence of microsegregation on the precipitation initiation time cannot be ignored, and so EPMA analysis was employed to assess the distribution of elements (Cr and Mo) for both weld metals. Fig. 16 shows the element profiles of Cr and Mo for the 316FR AF and FA weld metals. Cr and Mo are enriched at the cell or dendrite boundaries where δ-ferrite exists in the 316FR AF weld metal. Although Cr and Mo are also enriched at the cell boundaries (δ-ferrite areas) for the 316FR FA weld metal, the ferrite concentration differs. The concentration distributions were quantitatively compared by point analysis using EPMA. Table 7 shows the measured results for Cr, Mo, Fe, and Ni. All the values in both weld metals agree well with previously reported segregation behaviors for AF and FA mode weld metals [9,47,48]. Simply, segregation of the alloying elements is more severe in the AF mode (316FR AF weld metal) than that in the FA mode (316FR
Table 7 – EPMA analysis results (Point analysis, wt.%). Materials
Phases
Cr
Mo
Ni
Fe
316FR AF weld metal
δ-Ferrite Austenite δ-Ferrite Austenite
25.72 18.17 24.61 19.58
6.78 1.92 4.35 2.41
8.26 12.14 7.63 13.59
61.82 69.37 64.97 66.25
316FR FA weld metal
FA weld metal). Fig. 17 shows the Cr and Mo concentrations in the ferrite and the austenite. One outstanding feature is the concentration difference of Cr and Mo in δ-ferrite; ferrite in the 316FR FA weld metal contains less Cr and Mo than those in the 316FR AF weld metal, despite the fact that the 316FR FA base metal contains more Cr and Mo. Thus, the different precipitation initiation times in the σ + χ phase (Fig. 11) can be explained by the concentrations of Cr and Mo in the ferrite of the weld metal; that is, the rapid precipitation initiation in the 316FR AF weld metal is attributable to the higher concentrations of Cr and Mo in δ-ferrite compared to the 316FR FA weld metal. However, the precipitation speed of the 316FR FA weld metal becomes faster than that of the 316FR AF weld metal, and eventually the rate is reversed. This behavior is attributed to і) the different concentrations of Cr and Mo in the γ vicinity of the ferrite and іі) variation in the nucleation sites of the precipitates (amount of ferrite). In the as-welded situation, the Cr and Mo contents in δ-ferrite of the 316FR FA weld metal are smaller than those in the 316FR AF weld metal. During the
Fig. 16 – Microsegregation behavior of the alloying elements using EPMA for the 316FR AF and FA weld metals.
164
M A TE RI A L S CH A RACT ER IZ A TI O N 86 (2 0 1 3 ) 1 5 2–1 6 6
Fig. 17 – Point analysis of the microsegregation behavior using EPMA for the 316FR AF and FA weld metals.
aging treatment, Cr and Mo must diffuse from γ into δ-ferrite because the diffusion rates of Cr and Mo in ferrite are about 100 times faster than those in γ [49]. Fig. 17 also shows Cr and Mo concentrations in γ for both weld metals. γ for the 316FR FA weld metal has a higher concentration than that in the 316FR AF weld metal. Despite the initially slower precipitation initiation of the 316FR FA weld metal, diffusion of Cr and Mo from γ into δ-ferrite must occur as precipitation progresses because the 316FR FA weld metal, which possesses more Cr and Mo in γ provides more opportunities for precipitation events. Moreover, the available nucleation sites of the σ and χ phases (e.g., the ferrite area of the 316FR FA weld metal) are greater than those in the AF weld metal. Consequently, the rapid precipitation speed in the 316FR FA weld metal is due to і) large concentrations of Cr and Mo in γ and іі) the large amount of ferrite. Furthermore, the kinetic parameters (larger values for k0 and n in the 316FR FA than those of the 316FR AF weld metal) (Table 4) are
consistent with the rapid precipitation speed in the 316FR FA weld metal.
8. Prediction of the Aging Behavior Between 773 and 823 K for FBR Service Based on the determined kinetic equations for both weld metals, the decomposition behavior of δ-ferrite and the precipitation (σ + χ phases) behavior were predicted under the service conditions of a FBR (773 and 823 K) (Fig. 18). Initially, the 316FR FA weld metal has a lower decomposition rate of δ-ferrite than that of the 316FR AF weld metal, but the rate of the 316FR FA weld metal eventually exceeds that of the 316FR AF weld metal as operations progress. Table 8 lists the decomposition values of both weld materials at select aging times. Both weld metals have decomposition rates for δ-ferrite at an 823 K service temperature that are about 15 times faster than that at 773 K.
Fig. 18 – Prediction results: δ-ferrite decomposition ratio and σ + χ phase precipitation at the service temperatures of a FBR (773 and 823 K).
165
M A TE RI A L S C HA RACT ER I ZA TI O N 86 ( 20 1 3 ) 1 5 2–1 6 6
Fig. 18 and Table 9 show the prediction results for the σ + χ precipitation. Initially, the amount of σ + χ phases is similar, regardless of service temperature and weld metal. However, after 0.2 years of service, the amount of the σ + χ phases in the 316FR FA weld metal suddenly increases compared to that in the 316FR AF weld metal.
9.
Conclusions
Herein the microstructural changes and kinetic approaches of precipitates (σ and χ) during aging treatment for two types of weld metals of 316FR stainless steel (316FR AF and FA) were examined. The main conclusions are summarized below: (1) The 316FR weld metals contain approx. 2% (5%) of δ-ferrite formed during the AF (FA) solidification mode. The main precipitates in δ-ferrite are σ, χ, and γ2, which nucleate at the inside or surrounding δ-ferrite during aging treatments at 823 to 1073 K. These precipitates consume all the ferrite for both weld metals. (2) The decomposition behavior of δ-ferrite (δ-ferrite decomposition, σ + χ precipitation, σ precipitation) in the 316FR AF and FA weld metals was examined using three different kinetic approaches: the time law, Johnson– Mehl, and Austin–Rickett equations. The Johnson– Mehl-type equation is the most applicable to the decomposition behavior of the ferrite. The precipitation behaviors of σ and χ phases can be controlled by the diffusion of main alloying elements (Cr, Ni, and Mo) in δ-ferrite. (3) Precipitation (σ + χ) between both solidification modes differs due to the precipitation initiation time and precipitation speed. The different concentrations of Cr and Mo in δ-ferrite in the as-welded situation explain the difference in the precipitation initiation time, while the difference in the precipitation speed is presumed to be due to the varying concentrations of Cr and Mo in γ and the different amount of nucleation sites (i.e., amount of δ-ferrite). (4) Decomposition of δ-ferrite and σ + χ precipitation in the 316FR AF and FA weld metals were predicted at practical operating temperatures of a FBR (773 and 823 K) using the determined Johnson–Mehl-type equation. The predicted decomposition rate of δ-ferrite
Table 8 – Prediction results: decomposed fraction of δ-ferrite for select times in Fig. 18. Aging time (year)
0.1 1 5 10 50 100
Decomposed fraction of δ-ferrite 773 K
823 K
316FR AF
316FR FA
316FR AF
316FR FA
0.28 0.55 0.76 0.84 0.96 0.98
0.10 0.47 0.89 0.97 1.00 1.00
0.54 0.84 0.96 0.98 0.99 1.00
0.47 0.97 1.00 1.00 1.00 1.00
Table 9 – Prediction results: area fraction of σ + χ for select times in Fig. 18. Area fraction of σ + χ (%)
Aging time (year)
773 K
823 K
316FR AF
316FR FA
316FR AF
316FR FA
0.22 0.47 0.76 0.90 1.26 1.38
0.17 0.79 1.89 2.55 3.75 3.87
0.44 0.85 1.19 1.33 1.54 1.59
0.71 2.39 3.67 3.85 3.90 3.90
0.1 1 5 10 50 100
at an 823 K service temperature is about 15 times faster than that at 773 K for both weld metals. Furthermore, in the initial service stage, the amount of σ + χ phases is similar, regardless of service temperature for both weld metals, but after 0.2 years of operations, the amount of σ + χ phases in the 316FR FA weld metal drastically increases compared to the 316FR AF weld metal.
Acknowledgment The present study includes the result of the “Core R&D program for commercialization of the fast breeder reactor by utilizing Monju” entrusted to the University of Fukui by the Ministry of Education, Culture, Sports, Science and Technology of Japan (MEXT).
REFERENCES [1] Iida K, Asada Y, Okabayashi K, Nagata T. Construction codes developed for prototype FBR Monju. Nucl Eng Des 1987;98:283–8. [2] Iida K, Asada Y, Okabayashi K, Nagata T. Simplified analysis and design for elevated temperature components of Monju. Nucl Eng Des 1987;98:305–17. [3] Nakazawa T, Kimura H, Kimura K, Kaguchi H. Advanced type stainless steel 316FR for fast breeder reactor structures. J Mater Process Technol 2003;143:905–9. [4] Takahashi Y, Shibamoto H, Inoue K. Study on creep-fatigue life prediction methods for low-carbon nitrogen-controlled 316 stainless steel (316FR). Nucl Eng Des 2008;238:322–35. [5] Furukawa T, Kato S, Yoshida E. Compatibility of FBR materials with sodium. J Nucl Mater 2009;392:249–54. [6] Nogami S, Hasegawa A, Tanno T, Imasaki K, Abe K. High-temperature helium embrittlement of 316FR steel. J Nucl Sci Technol 2011;48:130–4. [7] Konomura M, Ichimiya M. Re-weldability tests of irradiated 316L(N) stainless steel using laser welding technique. J Nucl Mater 2002;307-11:1584–9. [8] Chun EJ, Baba H, Terashima K, Nishimoto K, Saida K. Prediction of σ phase embrittlement in type 316FR weld metal. Weld J May 2013:133–9. [9] Kou S. Welding metallurgy. 2nd ed. John Wiley & Sons Inc.; 2003 216–300. [10] Bae JI, Kim ST, Lee TH, Ha HY, Kim SJ, Park YH. High temperature precipitation behavior of high-nitrogen duplex stainless steel. Korean J Met Mater 2010;49:99–103.
166
M A TE RI A L S CH A RACT ER IZ A TI O N 86 (2 0 1 3 ) 1 5 2–1 6 6
[11] Byun SH, Kang N, Lee TH, Ahn SK, Lee HW, Chang WS , et al. Kinetics of Cr/Mo-rich precipitates formation for 25Cr–6.9Ni–3.8Mo–0.3N super duplex stainless steel. Met Mater Int 2012;1:201–7. [12] Lee TH, Ha HY, Hwang B, Kim SJ. Isothermal decomposition of ferrite in a high-nitrogen, nickel-free duplex stainless steel. Metall Mater Trans A 2012;43A:822–32. [13] Nakao Y, Nishimoto K, Inoue S. Kinetics of the formation of austenite at constant temperature — study on the phase transformation in welded joints of duplex stainless steels (report 1). J Jpn Weld Soc 1981;50:514–20. [14] Wasnik DN, Dey GK, Kain V, Samajdar I. Precipitation stages in a 316L austenitic stainless steel. Scr Mater 2003;49:135–41. [15] Barcik J. The kinetics of σ-phase precipitation in AISI 310 and AISI 316 steels. Metall Trans A 1983;14:635–41. [16] Magnabosco R. Kinetics of sigma phase formation in a duplex stainless steel. Mater Res 2009;12:321–7. [17] Nilsson JO, Wilson A. Influence of isothermal phase transformations on toughness and pitting corrosion of super duplex stainless steel SAF 2507. Mater Sci Technol 1993;9:545–54. [18] Nilsson JO, Huhtala T, Jonsson P, Karlsson L, Wilson A. Structural stability of super duplex stainless steel weld metals and its dependence on tungsten and copper. Metall Mater Trans A 1996;27A:2196–208. [19] Lee YH, Kim KT, Lee YD, Kim KY. Effects of W substitution on σ and χ phase precipitation and toughness in duplex stainless steels. Mater Sci Technol 1998;14:757–64. [20] Kim SB, Paik KW, Kim YG. Effect of Mo substitution by W on high temperature embrittlement characteristics in duplex stainless steels. Mater Sci Eng A 1998;247:67–74. [21] Chen TH, Weng KL, Yang JR. The effect of high temperature exposure on the microstructural stability and toughness property in a 2205 duplex stainless steel. Mater Sci Eng A 2002;338:259–70. [22] Michalska J, Sozanska M. Qualitative and quantitative analysis of σ and χ phases in 2205 duplex stainless steel. Mater Charact 2006;56:355–62. [23] Kim YJ, Ugurlu O, Jiang C, Gleeson B. Microstructural evolution of secondary phase in the cast duplex stainless steels CD3MN and CD3MWCuN. Metall Mater Trans A 2007;38A:203–11. [24] Escriba DM, Materna-Morris E, Plaut RL, Padilha AF. Chi-phase precipitation in a duplex stainless steel. Mater Charact 2009;60:1214–9. [25] Huang CS, Shih CC. Effect of nitrogen and high temperature aging on σ phase precipitation of duplex stainless steel. Mater Sci Eng A 2005;402:66–75. [26] Pardal JM, Tavares SSM, Fonseca MDPC, Souza JAS, Vieira LM, Abreu HFGD. Deleterious phase precipitation on superduplex stainless steel UNS S32750 characterization by light optical and scanning electron microscopy. Mater Res 2010;13:401–7. [27] Calliari I, Straffelini G, Ramous E. Investigation of secondary phase effect on 2205 DSS fracture toughness. Mater Sci Technol 2010;26:81–6. [28] Calliari I, Pellizzari M, Zanellato M, Ramous E. The phase stability in Cr–Ni and Cr–Mn duplex stainless steels. J Mater Sci 2011;46:6916–24. [29] Kashiwar A, Vennela NP, Kamath SL, Khatirkar RK. Effect of solution annealing temperature on precipitation in 2205 duplex stainless steel. Mater Charact 2012;74:55–63. [30] Pohl M, Storz O, Glogowski T. Effect of intermetallic precipitation on the properties of duplex stainless steel. Mater Charact 2007;58:65–71.
[31] Ghosh SK, Mondal S. High temperature aging behaviour of a duplex stainless steel. Mater Charact 2008;59:1776–83. [32] Gill TPS, Vijayalakshmi M, Rodriguez P, Padmanabhan KA. On microstructure–property correlation of thermally aged type 316L stainless steel weld metal. Metall Trans A 1989;20:1115–24. [33] Gill TPS, Shankar V, Vijayalakshmi M, Rodriguez P. Influence of carbon on the transformation kinetics of delta-ferrite in type 316 stainless steel weld metals. Scr Metall Mater 1992;27:313–8. [34] Gill TPS, Shankar V, Pujar MG, Rodriguez P. Effect of composition on the transformation of δ to σ in type 316 stainless steel weld metals. Scr Metall Mater 1995;10:1595–600. [35] Mathew MD, Sasikala G, Mannan SL, Rodriguez P. Effect of stress on the transformation behaviour of delta-ferrite in type 316 stainless steel weld metal. J Mater Sci Lett 1994;13:194–6. [36] Sasikala G, Ray SK, Mannan SL. Kinetics of transformation of delta ferrite during creep in a type 316(N) stainless steel weld metal. Mater Sci Eng A 2003;359:86–90. [37] Chun EJ, Baba H, Terashima K, Nishimoto K, Saida K. Prediction of σ phase precipitation in type 316FR stainless steel weld metal. J Jpn Weld Soc 2013;31:168s–72s. [38] Badji R, Bouabdallah M, Bacroix B, Kahloun C, Berrahar K, Kherrouba N. Effect of solution temperature on the precipitation kinetic of σ-phase in 2205 duplex stainless steel welds. Mater Sci Eng A 2008;469:447–54. [39] Inoue H, Koseki T, Okita S, Fuji M. Solidification and transformation behavior of austenitic stainless steel weld metals solidified as primary austenite — study of solidification and subsequent transformation of Cr–Ni stainless steel weld metals (1st report). Weld Int 1997;11:876–87. [40] Inoue H, Koseki T, Okita S, Fuji M. Solidification and transformation behavior of austenitic stainless steel weld metals solidified as primary ferrite — study of solidification and subsequent transformation of Cr–Ni stainless steel weld metals (2nd report). Weld Int 1997;11:937–49. [41] Johnson WA, Mehl RF. Reaction kinetics in processes of nucleation and growth. Trans AIME 1939;135:416–59. [42] Starink MJ. Kinetic equations for diffusion-controlled precipitation reactions. J Mater Sci 1997;32:4061–70. [43] Nakao Y, Nishimoto K, Ohtani K. Transformation kinetics of δ ferrite during heat treatment. J Jpn Weld Soc 1984;2:61–7. [44] Lee ES, Kim YG. A transformation kinetic model and its application to Cu–Zn–Al shape memory alloys—1. Isothermal conditions. Acta Metall Mater 1990;38:1669–76. [45] Porter DA, Easterling KE, Sherif MY. Phase transformations in metals and alloys. 3rd ed. New York: CRC Press; 2009. [46] Reed-Hill RE, Ababschian R. Physical metallurgy principles. 3rd ed. Boston: PWS Publishing Company; 1992. [47] Brooks JA, Williams JC, Thompson AW. STEM analysis of primary austenite solidified stainless steel welds. Metall Trans A 1983;14A:1271–81. [48] Brooks JA, Williams JC, Thompson AW. Microstructural origin of the skeletal ferrite morphology of austenitic stainless steel welds. Metall Trans A 1983;14A:23–31. [49] Hsieh CC, Wu W. Discussing the precipitation behavior of σ phase using diffusion equation and thermodynamic simulation in dissimilar stainless steel. J Alloys Compd 2010;506:820–5.