In this study, we investigated the correlation between the microstructure and the localized corrosion characteristics for a weld metal made up of 22Cr-9Ni duplex stainless steel (DSS) aged in the range of 400–1000 °C for 30 min. The results showed that changes in the microstructure of the DSS weldments, owing to the formation of precipitates during the aging process, affected its pitting corrosion resistance. The microstructural evolution was found to be extremely sensitive to the secondary phase precipitations such as chi (χ), sigma (σ), chromium nitride (Cr2N), and secondary austenite (γ2) in the aging range of 700–900 °C. The potentiodynamic polarization tests confirmed that the pitting potential representing the corrosion resistance decreased gradually from 500 °C, attained the lowest at 850 °C, and increased again at 900 °C. In addition, the Cr or Mo deficient area around σ phase was transformed into the secondary austenite (γ2) phase during aging where the pitting was found to occur.
Duplex Stainless Steel (DSS), owing to its metallurgical structure comprising ferrite and austenite phases in equal proportions, exhibits high mechanical properties and corrosion resistance when compared with other types of stainless steel, and it is used in a variety of highly corrosive environment such as petrochemical, nuclear power, marine structure [1‐4], however, if it is exposed to a high temperature during a manufacturing process such as welding or heat treatment, the ferrite–austenite phase ratio changes along with the precipitation of secondary phase [5, 6].
The process of secondary phase precipitation in DSS is shown in the time–temperature-precipitation (TTP) diagram (Fig. 1 [7]), whereby the temperature scale is divided into two distinctly different. A low temperature section of 400–500 °C where Cr-rich α′ appears owing to the spinodal decomposition of ferrite, and a high temperature section of 600–1000 °C where the intermetallic compounds (chromium carbides and nitride) or precipitates (χ phase, σ phase, and γ2) coexist [8‐10]. As these precipitates deteriorate the mechanical characteristics and corrosion resistance, many research studies on the secondary phases have been in progress [11‐14]. Tavares et al. [15, 16], reported that, α′ phase is formed rapidly at 475 °C causing corrosion and brittleness in DSS. Zhang et al. [10] reported that a Cr deficient area around Cr2N formed through the heat treatment at 700 °C for 3–240 min was transformed into the secondary austenite phase (γ2), which deteriorated the corrosion resistance. In addition, Deng et al. [17] reported that a pronounced σ phase with fast formation kinetic was formed at 850 °C, which improved the intensity and decreased the corrosion resistance significantly.
Fig. 1
Schematic diagram for precipitate formation in duplex stainless steels. (Redrawn from Charles [7])
×
Advertisement
Even though there are many studies on the aging behavior of DSS, the relationship between the microstructure and the corrosion characteristics according to the formation of the secondary precipitates has not yet been reported in detail. The purpose of the present work was to investigate the effect of aging, especially at high temperatures, on the microstructural evolution and the pitting potential in 2209 DSS weldments. This work focused specifically on the relationship between the microstructure and the pitting corrosion resistance for the DSS aged in the range of 400–1000 °C [18‐22].
2 Experimental Details
The specimens of DSS weld metal with a dimension of 20 mm × 15 mm used in this investigation were subjected to the Flux Cored Arc Welding (FCAW) process with 100% CO2 shielding gas and ER 2209 filler wires and their chemical composition is shown in Table 1.
Table 1
Chemical composition of the weld metals (wt%)
C
N
Si
Mn
P
S
Cr
Ni
Mo
Cu
Fe
FN
As-welded metal
0.029
0.114
0.82
0.59
0.025
0.008
22.03
8.73
3.28
0.07
Bal.
50–60
The specimens were subjected to an isothermal heat treatment in the aging range of 400–1000 °C for a nominal duration of 30 min, followed by quenching in water. The volumetric percentages of ferrite and austenite phases were estimated using a ferritescope. The as-welded metal Ferrite Number (FN) obtained using this method is shown in Table 1. The metallographic sections that are transverse to the welding direction were prepared using standard techniques of mechanical polishing. The specimens were then etched electrolytically using a 20% KOH electrolyte at 3 V etching potential for 20 s. Before and after the anodic polarization test, the microstructures of the specimens were examined using an optical microscope and a scanning electron microscope (JSM-6700F, JEOL). An energy dispersive X-ray spectrometer (EDS) attached to the SEM was used to verify the existence of the secondary phases at an accelerating voltage of 40 kV and a spot size of 4.0. The measurement was repeated five times in order to reduce the margin of error. Foils with a thickness of 0.05 mm, prepared by way of abrasion on SiC papers and electropolishing (twin-jet) in a mixture of 5% perchloric acid and 95% ethanol at − 20 °C and 40 V using a Struers Tenupol twin-jet unit, were examined using JEOL 2100 TEM operated at 200 kV.
All the electrochemical measurements were carried out using a potentiostat Versa STAT3 (AMETEK), and the electrochemical cell was made of a glass beaker suspended in a water bath with three electrodes (reference, counter, and working) immersed directly into the test solution. A silver/silver chloride (3 mol L−1 KCl, 0.197 V) was normally used as the reference electrode (all potentials refer to this scale), a specimen used as the working electrode, and a flat coil of platinum used as the counter electrode [24, 25]. The specimens acting as working electrodes were embedded in epoxy resin. Prior to each experiment, the working electrode was ground by using a 1000-grit emery paper, cleaned ultrasonically with ethanol, rinsed with distilled water, and dried in air. In order to avoid crevice corrosion, the interfaces between the specimen and the resin were sealed with special polyacrylate resin (non-polarity) and dried in the air. The test solution of 1 mol L−1 NaCl was made up of an analytical grade reagent and distilled water. Before the potentiodynamic polarization test, the working electrodes were immersed in the electrolytes for at least 20 min for stabilization of the open-circuit potential (OCP). The polarization measurement was conducted at 25 ± 1 °C respectively at a scan rate of 0.4 mV/s starting from − 500 to 600 mV. The pitting potential (Epit) was identified as the potential at which the current density sharply increased. The test solution was bubbled using pure nitrogen gas (N2) to prevent entry of oxygen gas (O2) throughout the test. All tests were carried out at least three times to ensure reliability and repeatability. It was shown experimentally that this scan rate had little effect on the shape of polarization curves. Any crevice corrosion observed on the specimen after the testing was discarded considering that the test results were invalid.
Advertisement
3 Results and Discussion
3.1 Microstructure Evolution During Aging
Figure 2 shows the microstructure of an as-welded specimen and the microstructure of a weld metal specimen which was subjected to an isothermal transformation treatment for 30 min at temperatures in the range of 400–1000 °C, respectively. Figure 2a shows the typical multi-pass welding area, composed of white grain boundary or Widmannstätten austenite at gray ferrite matrix. On the contrary, carbides, which are easily formed by welding cycles, are not identified in the microstructure, and it is believed that the carbon contents amounted to less than 0.03 wt%. Several papers reported that the DSS would be decomposed through into chromium-depleted α and chromium-rich α′ owing to the spinodal decomposition of ferrite [15, 16, 20], when aged at 350–550 °C for more than 100 h. However, as shown in Fig. 2b–d, there may not be any significant change in the microstructure even after the heat treatment for 30 min at 400–600 °C. After aging at 700–900 °C, the intermetallic phases such as sigma (σ) phase and chi (χ) phase were precipitated in the structure, and as shown in Figs. 2e–g and 3, it is found that they are formed in the grain boundary of δ/γ phase in the initial stage and grown toward the α phase subsequently. At 1000 °C, the secondary precipitates were reduced relatively. The progressive variation of the ferrite volume fraction during the isothermal heat treatment was measured using a ferrite scope, and the results are shown in Fig. 4. As the aging temperature increases, the amount of ferrite decreases and the lowest content of 1.9 FN are observed at 850 °C. As the secondary precipitates are reinforced in the matrix from 900 °C onwards, the ferrite phase increases gradually above 950 °C. This phenomenon could be explained as the transformation kinetics of several precipitations according to the aging temperature. The precipitation of the secondary phase is found to be the most active phase at 850 °C which is eventually redissolved to ferrite from 900 °C onwards. As a result, the meta-stable δ-ferrite below 1000 °C would ultimately be decomposed into Cr, Mo-enriched σ phase and Cr, Mo-depleted γ phase as shown Fig. 1 [7, 23].
Fig. 2
Optical microstructure obtained from duplex stainless steel weldments aged at different temperature of a as-weld, b 400 °C, c 500 °C, d 600 °C, e 700 °C, f 750 °C, g 850 °C and h 1000 °C for 30 min
Fig. 3
Transmission electron micrograph obtained from specimen aged at 850 °C for 30 min: a showing sigma (σ) phase grown toward δ-ferrite region. b Schematic representation of the mechanism for preferential growth direction of sigma phase
Fig. 4
Ferrite number of specimens aged at different temperature for 30 min
×
×
×
3.2 X-ray Diffraction Results
Figure 5 shows the results of X-ray diffraction of the specimens that were subjected to an aging treatment for 30 min at 500 °C, 700 °C, 800 °C, and 1000 °C, respectively. At 500 °C, and 700 °C, only the main phases, ferrite (δ) and austenite (γ), are seen, however, this does not mean that the secondary phases do not exist, and it is expected that either their concentration is less than 3% or they exist as clusters of atom that is typical of the initial stage of precipitates with a capability to form the coherent relation with the matrix. Up to 800 °C, it shows the peaks of σ phase strongly, while it shows that the intensity of σ phase decreases gradually from 900 °C. At 1000 °C, most of the secondary phases are redissolved as shown in Fig. 1 and the fraction of ferrite increases.
Fig. 5
XRD analysis results of the specimens: a as-weld, b 500 °C, c 700 °C, d 800 °C, e 1000 °C
×
3.3 Effect of Different Aging Temperature on Pitting Corrosion Resistance
To evaluate the corrosion resistance characteristics of the DSS welding area by the heat treatment temperature, the potentiodynamic polarization test was carried out in an aqueous solution of 25 ± 1 °C 1 mol L−1NaCl. The results are shown in Fig. 6a, and the measured pitting potentials are shown in Fig. 6b. The specimen that was subjected to the isothermal aging treatment was measured three times under the same condition, and the average value of the measured data was considered. It appears from these figures that the aging temperature affected the passivation properties of the weld metal. The pitting potential of as-welded specimen has the highest value at 504 mV(Ag/AgCl2) and the measured passivation area is found to be the largest. The polarization curve of the specimen that was subjected to a thermal treatment at 400 °C for 30 min shows that the current density increases marginally compared to that of the as-welded specimen, while the pitting potential is found to be the same value of 500 mV(Ag/AgCl2). This means that Cr-rich α′ phase is not formed by the aging for 30 min at 400 °C, and it does not affect the pitting corrosion resistance at all. At 500 °C, it shows 390 mV(Ag/AgCl2) which means that the pitting potential has decreased relatively, attributable to the spinodal decomposition of ferrite. As the temperature increases up to 800 °C, the passivity current density shows an incremental change by 0.1µA/cm2, while the pitting potential has decreased gradually, and at 850 °C, it shows the lowest pitting potential of 90 mV(Ag/AgCl2). As the pitting potential decreases, the susceptibility to pitting corrosion becomes higher, and as the current density increases, more current flows, thereby increasing the corrosion rate. From 900 °C, the passivity current density shows an increase by 1 µA/cm2, and the pitting potential shows a small increase. The pitting potential is 410 mV(Ag/AgCl2) at 1000 °C. This phenomenon is attributed to the decomposition of secondary precipitates that gets highly activated and reinforced in the matrix for a temperature above 900 °C compared with the DSS diagram for precipitation in Fig. 1 and the ferrite number graph shown in Fig. 4.
Fig. 6
The results of corrosion resistance obtained from the specimens aged at different temperature for 30 min in a 1 mol L−1 NaCl solution: a the polarization curves, b pitting potential curves
×
3.4 The Characteristics of Pit Morphology Based on Precipitation Types
After the potentiodynamic polarization tests, the pits formed on the surface of each specimen was observed. In the case of the as-welded specimen, a metastable pit is formed at δ/γ phase and propagated to the direction of ferrite while for the aging-treated specimen, the pitting traces were observed around the secondary precipitates. Figure 7 shows the pitting morphology of the specimen that had a heat treatment for 30 min at 400 °C. While the metastable pit occurred primarily in the surrounding vicinity of the secondary austenite, the stable pit in the ferrite matrix occurred in ① area. As observed from the EDX analysis, compared with the matrix of ferrite and austenite, the Cr content decreased to 19.21 wt%, whereas Mo (13.71 wt%) and Si (2.98 wt%) increased sharply, Cr, Mo and Ni decreased sharply and the corrosion occurred preferentially in the area surrounding these precipitates (② area). According to Iacoviello et al. [24], in the case of tempering of DSS at 300–500 °C, Ni, Si, Mo-rich g-phase similar to Cr-rich α′ phase are formed in dislocation or δ/γ interfaces, which becomes the main cause of localized corrosion.
Fig. 7
G phase precipitation and spinodal decomposition occurring in the ferrite phase after aging treatment, identified by SEM
×
At 600–700 °C, as shown in Fig. 8, a small amount of Mo-rich Laves (Fe2Mo) phase was precipitated in intragranular or intergranular sites, which decreased the pitting corrosion resistance around those sites. At more than 700 °C, chi (χ) phase, sigma (σ) phase, chromium nitrides and secondary austenite (γ2) phase were formed to facilitate reduction in the corrosion resistance of welding area. [2, 11, 25, 26].
Fig. 8
SEM morphology of the corrosion attack of the specimen aged at 600 °C for 30 min (Laves phase)
×
Figure 9a shows the regions where the secondary phases are precipitated, in the case of the electrolytically etched specimens. The chemical compositions of the existing phases have been evaluated using EDX, and the results obtained are presented in Fig. 10 and Table 2. Figure 9b, c show the finish of the microscopic structures at the instant when the corrosion test stopped and the polarization curve displayed the passivation area during the corrosion experiment on the specimen that had an aging treatment for 30 min at 750 °C. Compared with the ferrite number and the trend of the precipitates according to the heat treatment shown in Fig. 4, it is found that these precipitates occurred by consuming the ferrites. Chen et al. [26] also revealed in his paper that the chemical composition of intermetallic phase such as χ phase and σ phase is close to the ferrite phase and that the main elements (Cr, Mo) which form these precipitates diffuse in the ferrite rapidly. As shown in Fig. 9b, c, χ phase is formed like a bright and thin line in δ/δ subgrain or δ/γ interface, and it shows the trend of progressive corrosion by σ phase. This matches with the fact that a normalized thermodynamic driving force of χ phase is larger than that of σ phase at 700–900 °C section, and according to Karlsson [27], the nucleus is formed faster than the σ phase due to the low lattice coherency stresses of χ phase. At temperatures more than 800 °C, the precipitates are formed on σ phase more than on χ phase, which means that the χ phase in the metastable stage increases the formation speed at more than 800 °C so that it is formed within 30 min.
Fig. 9
Scanning electron microscopy micrographs of sigma, chi and new austenite phase in the DSS weldment with the aged condition: a 750 °C for 30 min followed by quenching in water. etching: electrolytic 30% KOH at 6 V 20-30 s, b and c SEM morphology of the corrosion attack on the same specimen after anodic polarization test, d bright field TEM micrograph of χ and σ precipitates
Fig. 10
EDX spectra at a chi, b sigma and c secondary austenite phase, as marked in Fig. 9a
Table 2
Major alloy contents measured at the marked regions in Fig. 9a (wt%)
Cr
Ni
Mo
N
Mn
Si
σ-phase
29.52
5.28
6.81
0.43
0.36
1.52
χ-phase
23.02
4.72
10.72
0.37
0.44
1.42
ferrite
24.77
7.55
4.23
0.12
0.27
1.23
γnew
19.47
9.90
1.80
0.41
0.51
0.80
×
×
In general, ferrite is transformed into σ phase and γ2 phase by the eutectoid type reaction, and it is well known that σ phase is formed in δ/γ interface and grown in intragranular site of ferrite [28, 29]. Figure 11 shows σ phase formed in δ/γ interface and the pitting occurrence in the surrounding area (intragranular γ2, intergranular γ2) preferentially. Besides this, micro chromium nitrides of 0.5–0.8 µm were observed in δ/γ interface (Fig. 12), and according to Ramirez [30], this also acts as the driving force to form γ2 phase. Thus, it is found that when the secondary austenite is exposed to the actual corrosion environment, it provides sites for the pit initiation. In addition, it shows that the pit size decreases and localized locally in certain areas above 1000 °C (Fig. 13), the precipitates is accelerated and the pitting potential increases sharply.
Fig. 11
The mechanism of pitting corrosion for the specimen aged at 850 °C for 30 min: a the nucleation and growth of σ phase at δ/γ interface depending on the diffusion of Cr and Mo, b and c pitting morphologies at Cr-depletion area around the σ phase, i.e. in the newly formed secondary austenite phase, d Bright field TEM micrograph of σ precipitates
Fig. 12
Nitrides precipitated at the δ/γ interfaces in the specimen aged at 900 °C for 30 min: a intergranular nitride morphologies. SEM image, b detail of the precipitates. Bright-field TEM image
Fig. 13
SEM morphology of the corrosion attack of the specimen aged at 1000 °C for 30 min
×
×
×
4 Conclusions
To analyze systematically the corrosion resistance as well as the microstructure of the DSS welds of 22 Cr-9 Ni-3.3Mo (wt%) according to its aging temperature, a potentiodynamic polarization test was carried out after an isothermal treatment for 30 min at 400–1000 °C followed by water quenching, and the corrosion morphology was observed using SEM and TEM. Conclusions from these studies are as follows:
(1)
There was no remarkable change in the ferrite (FN) content and the optical images, up to 600 °C. As the aging temperature increased, FN decreased and a number of secondary phases were precipitated in intergranular sites of δ/γ and ferrite. Above 900 °C, the precipitates were reinforced and FN increased again, indicating that the precipitates have been redissolved and they can affect the corrosion resistance.
(2)
The anodic polarization tests, revealed that the as-welded specimen showed the highest corrosion resistance of 504 mV (Ag/AgCl2), up to 400 °C, as the temperature increased, the corrosion resistance decreased gradually, and the specimen corresponding to 850 °C showed the lowest value of 90 mV (Ag/AgCl2). Above 900 °C, it began to increase again drastically up to 410 mV (Ag/AgCl2) at 1000 °C.
(3)
The pitting morphology analysis indicated that the pit trace of tire-track type appeared at 500 °C. From TEM analysis, it was evident that this pit trace appeared because the α′ phase which occurred owing to the spinodal decomposition of ferrite was laminated in a line causing local corrosion in Cr-depleted region. Above 700 °C, χ phase and σ phase were generated in δ/δ or δ/γ interfaces to form the secondary austenite (γ2) in intragranular sites of ferrite or intergranular sites of the first austenite. From 850 °C onwards, χ phase decreased sharply and σ phase appeared mainly, and above 900 °C, due to the reinforcement of secondary precipitates, σ phase began to decrease inside the deposited area, and at 1000 °C, relatively small-sized σ phase cluster was observed.
(4)
In conclusion, at the heat treatment section of 500–850 °C, most precipitates were formed, thus the Cr and Mo deficient regions occurred around the precipitates and acted as a pathway where the dissolution happened in the corrosion environment preferentially.
Acknowledgements
This study was supported by the Dong-A university research fund.
