Corrosion resistance of super duplex stainless steels in chloride ion containing environments: investigations by means of a new microelectrochemical method: II. Influence of precipitates
Introduction
Super duplex stainless steels (SDSS) are being increasingly used as structural materials in marine and petrochemical applications. This is mainly due to their high resistance to corrosion and stress corrosion cracking as well as to their high strength and toughness.
In addition to ferrite and austenite, a large variety of undesired secondary phases may be formed in SDSS during isothermal ageing or quenching. The undesirable formation of secondary phases reduces the overall corrosion resistance and makes the alloy more susceptible to localised corrosion attack. Moreover, such precipitations may also affect the mechanical properties, mainly the toughness.
The chemical composition and the heat treatment temperature have the most important influence on the precipitation behaviour of SDSS. Both determine the phase volume fraction of ferrite and austenite and the partitioning of the main alloying elements, chromium, molybdenum, nickel and nitrogen [1].
A companion paper (Part I) deals primarily with “precipitation-free” materials [2] and is directly connected to this present Part II. Here, the influence of precipitates is determined.
In Section 1.1, a short description of the investigated secondary phases (sigma phase, chromium nitride and secondary austenite) is presented. It must be emphasised that the technically relevant states considered here are not necessarily in thermodynamic equilibrium, and that the state of the secondary phase depends on the previously applied heat treatment.
Sigma phase is by far the most important secondary phase because of its relatively large volume fraction and its detrimental influence on toughness and corrosion resistance. Especially in SDSS, the high contents of chromium and molybdenum, the main promoting elements for sigma phase formation, facilitate the precipitation of the sigma phase [3]. In practice, when the annealing times are generally too short to approach thermodynamic equilibrium, there are different ways of obtaining sigma phase. Sigma phase can be formed by passing through the temperature range of the thermodynamical stability (“C” curve of the sigma phase), either by slow heating or by slow cooling. Slow heating through this temperature range leads to the formation of sigma phase. Going on to slightly higher annealing temperatures than this temperature range will not cause the sigma phase to disappear after reasonably long annealing times (30 min). Slow cooling from high temperatures down to the temperature range where the sigma phase is stable also results in the formation of sigma phase, but at lower temperature than sigma phase is found in the first case. Sigma phase is often found to precipitate at triple junctions where two austenite grains meet one ferrite region, and the sigma phase tends to grow into the ferrite phase corresponding to the eutectoid reaction δ→σ+γsec; i.e. the former ferrite is transformed into sigma phase and secondary austenite (Fig. 1) [4]. The growing sigma phases will consume chromium and molybdenum not only from the ferrite, but also from the primary austenite [5].
In most cases sigma phase formation can be suppressed by the appropriate thermal treatment. Annealing at high temperatures increases the ferrite content and dilutes the ferrite with respect to the sigma phase promoting elements. Also the less pronounced partitioning at higher temperatures enhances this effect. Subsequent quenching from high annealing temperatures prevents the formation of sigma phase, but favours the formation of chromium nitrides.
Rapid cooling (water quenching) from high solution temperatures causes a supersaturation of nitrogen in the ferrite and leads to the formation of chromium nitride [6]. Chromium nitrides of this type nucleate at single dislocation lines. They are accumulated mainly in the centre of ferritic regions and rarely found close to the ferrite–austenite boundary (Fig. 2): during quenching, nitrogen close to the phase boundary may escape into the adjacent austenite phase which provides a much higher solubility for nitrogen. The frequently observed accumulations and alignment of chromium nitrides at ferrite subgrain boundaries correspond to the high dislocation density of those boundaries and can be explained through the diffusion impediment of nitrogen at these dislocations during quenching.
Chromium nitrides may also be formed in the ferrite during isothermal heat treatment at lower temperatures. These nitrides, generated in an equilibrium reaction, are coarser plates and equally distributed at the ferrite subgrain boundaries and dislocation nodes. However, by means of light optical microscopy it is difficult to distinguish between equilibrium nitrides and nitrides formed during quenching.
In the current work, only the chromium nitrides generated by fast quenching from high solution annealing temperatures will be considered for the measurements of the corrosion resistance. The influence of chromium nitrides formed by isothermal ageing at lower temperatures was not investigated.
Secondary austenite in the shape of Widmannstätten needles in the ferrite region (Fig. 3) is formed by heating up slowly through a temperature range around 1000°C to annealing temperatures below 1150°C [7]. The temperature range depends on the alloy composition.
Quenching from high solution annealing temperatures as 1300°C can also result in the formation of secondary austenite in the shape of Widmannstätten needles in ferrite phase regions together with the formation of chromium nitrides. These secondary austenite needles generated by quenching are not subject of the present investigation, only the Widmannstätten type of secondary austenite formed during heating is considered here.
Section snippets
Experimental procedure
The experimental procedure such as melting procedure and heat treatment, is the same as the procedure described in Part I. The peculiarities concerning the precipitates will be mentioned here.
Metallography
The precipitation behaviour for the different heat treatment of the investigated alloys is summarised in Table 2. The classification of the amounts of precipitates is subjective, since it is almost impossible to evaluate small precipitates, e.g. chromium nitrides, quantitatively. From Table 2 it is obvious that no really precipitation free state can be achieved after quenching from solution annealing temperatures in the range of 900–1300°C [2]. This seems to be a big disadvantage of the
Discussion
It must be emphasised that the nitrogen content of an alloy is a crucial factor with respect to its corrosion resistance because nitrogen contributes strongly to the empirical PREN. The conclusions on the corrosion behaviour are based principally on the elements chromium and molybdenum and not on nitrogen.
The eutectoid decomposition of ferrite into sigma phase and secondary austenite has the most detrimental effect on the corrosion resistance. The weakest point in corrosion resistance is the
Conclusions
(1) Sigma phase formation in SDSS is connected with the formation of eutectoid secondary austenite and a depleted zone in the primary austenite – both poor in chromium and molybdenum – and leads to a huge decrease in corrosion resistance.
(2) Lower sigma phase formation temperatures (until the tip of the “C” curve of the sigma phase) results in an higher volume fraction of sigma phase and a weaker secondary austenite with respect to the corrosion resistance. The corrosion resistance is further
Acknowledgements
The authors wish to thank Edi Schaller from the Institute of Metallurgy, Swiss Federal Institute of Technology, Zürich, Switzerland for his support in metallography.
References (17)
- et al.
Mater. Sci. Engng. A
(1998) - et al.
Calphad Computer Coupling of Phase Diagrams and Thermochemistry
(1985) Calphad Computer Coupling of Phase Diagrams and Thermochemistry
(1993)- et al.
Electrochim. Acta
(1995) - R.A. Perren, P.J. Uggowitzer, M.O. Speidel, Proceeding of the Duplex Stainless Steels World Conference, Maastricht,...
- et al.
Metal. Sci.
(1983) Mater Sci Technol
(1992)- J.-O. Nilsson, Proceedings of the Duplex Stainless Steels World Conference, Maastricht, 1997, pp....
Cited by (124)
Unexpected failure of cast superduplex stainless steel exposed to high chlorides containing water: From failure analysis to corrosion mechanisms settlement
2022, Engineering Failure AnalysisCitation Excerpt :This can induce a different distribution of ferrite and austenite stabilizing elements in each phase, leading to a local variations of the PREN number possibly affecting the corrosion properties at the phase with lower PREN [4–8]. Corrosion properties could become even worse if sigma phase or Cr nitrides are formed during thermal treatments [2–3,9]. Corrosion behavior of super duplex stainless steels is strongly affected by the aggressiveness of the environment (in particular Cl− and sulfur content) [6,8–10,12–13] and by the service temperature [7,10–11].
Investigation and optimization of turning process parameters in super duplex stainless steel
2022, Materials Today: ProceedingsApplication of bipolar electrochemistry to assess the ambient temperature corrosion resistance of solution annealed type 2205 duplex stainless steel
2022, Materials Chemistry and PhysicsCitation Excerpt :However, for lean DSS 2101 (21%Cr-1%Ni), an optimized PREN between ferrite and austenite cannot be achieved by heat treatments [20]. Lower temperature annealing treatment might precipitate other phases, such as sigma phase, chi phase, and chromium nitride, which all reduce the corrosion resistance [21–23]. Heat treatment applied to duplex stainless steel can also affect stress corrosion cracking behaviour.
Measuring changes in wettability and surface area during micro droplet corrosion measurements
2021, Electrochimica ActaEffects of reducing atmospheres of bright annealing on the surface and corrosion characteristics of super duplex stainless steel tubes
2021, Surface and Coatings TechnologyCitation Excerpt :Compared to the case of N15, N30 showed higher concentrations of both N and Cr, in the matrix below the oxide layer, suggesting that Cr2N was precipitated in the surface region of N30. From general corrosion theory, the local depletion of Cr [19,49], adjacent to the Cr2N precipitates in the sample, weakens the stability of the oxide scale, making it vulnerable to localized corrosion. On the other hand, as shown in Fig. 9(c) and (d), there was no significant difference in the thickness of the oxide scale formed on the two samples (H1050 and H1100).