**2. General Characteristics of Duplex Steel**

Duplex steels can be divided into five basic categories, depending on the percentage of alloying elements (Cr, Mo, Ni, Mn, Cu, and N—Figure 1). These groups are as follows:


**Figure 1.** Chemical compositions and Pitting Resistance Equivalent Numbers (PREN) for various categories of duplex steels—according to [21].

A typical microstructure of properly balanced duplex stainless steel is given in Figure 2. It includes ferritice (dark) and austeniteic grains (light). The chemical composition of duplex stainless steel is limited by thermodynamic stability of austenite and ferrite and also by nitrogen solubility limit. Stability areas of duplex steel with respect to combined Cr + Mo mass fraction are given in Figure 3. Below 20% Cr + Mo, there is a risk of martensitic transformation of austenite. At above 35% Cr + Mo a δ-ferrite instability and formation of harmful secondary phases may occur. Moreover, a variation of nitrogen content in HDSS may influence the phase stability. Such difficulties are, in general, caused by high nitrogen vapor pressure.

**Figure 2.** A properly balanced duplex steel microstructure containing approximately 50% of each δ-ferrite (dark areas) and γ-austenite (light areas)—according to [21].

**Figure 3.** Stability areas of conventional duplex steels—according to [21].

Balanced duplex stainless steels are theoretically located at a 50% ferrite content line in Schaeffler–DeLong diagrams (Figure 4). However, in practice in properly balanced steel microstructure, the ferrite—austenite proportions range are wider, i.e., ferrite austenite <sup>≈</sup> 50%−10% 50%+10% .

**Figure 4.** Location of individual categories of duplex steels in the Schaeffler–DeLong diagram according to [21].

The appropriate use of duplex steel demands the welding engineers perform thoughtful actions based on well-established engineering knowledge and experience. A commonly applied routine-based approach to welding admittedly leads to obtaining the expected mechanical properties of welding joints. Nevertheless, it does not guarantee the concurrent obtaining of required corrosion resistance for the joints. It is known that the corrosion resistance of joints in the chloride environment impact zone reaches only 50–80% of the parent material corrosion resistance [22].

#### **3. PREN Chloride Corrosion Resistance Index**

In the right of Figure 1, the values of Pitting Resistance Equivalent Number (PREN) have been given for each group of duplex steels. This index is used to evaluate the pitting corrosion resistance of the steels in a chloride environment. The PREN indicator refers to the thermodynamically stable steels, i.e., the steels after their final heat treatment [23]. In the case of duplex steels, the following formula given by Herbsleb is used to calculate the value [24]:

$$\text{PREN} = \text{Cr} + \text{3.3Mo} + 16\text{N} \tag{1}$$

In this equation Cr, Mo, and N are weight percentages of the corresponding elements. For SDSS and HDSS, which contain W or Cu, different formulae may be used [24]. These are as follows:

• Okamoto formula:

$$\text{PREINW} = \text{Cr} + 3.3(\text{Mo} + 0.5\text{W}) + 16\text{N} \tag{2}$$

• Heimgartner formula:

$$\text{PRENCu} = \text{Cr} + 3.3\text{Mo} + 15\text{N} + 2\text{Cu} \tag{3}$$

• Extended formula:

$$\text{PRENEXT} = \text{Cr} + 3.3(\text{Mo} + 0.5\text{W}) + 2\text{Cu} + 16\text{N} \tag{4}$$

In these formulae, the content of a particular alloying element is given in its mass percentage (%wt.).

The PREN values range from 26 (for LDSS steels with average pitting corrosion resistance) up to above 45 (for HDSS steels with high corrosion resistance). In both cases, the resistance is significantly higher when compared to conventional steel grades. However, the values of PREN should be treated as comparative data. The final selection of steel for a given application should be based on tests carried out in a given corrosive medium. The usefulness of the PREN indicator for estimating the analogous resistance of welding consumables is limited due to different welding techniques that can be used and hence the different levels of nitrogen introduction into the welded melt metal. The terms "high corrosion resistance" and "average corrosion resistance" should be therefore interpreted as relative measures.

#### **4. Alloying Elements Influence on Duplex Steels Corrosion Resistance**

Duplex steels usually crystallize from a liquid in the form of δ-ferrite. During alloy cooling, the lattice structure stresses increase because of Fe replacement by elements with larger radii, e.g., Ni.

In the areas with a higher concentration of austenite-forming elements, the A2-type ferrite lattice structure is transformed to an A1-type lattice (with 25% larger lattice parameter) at δ-solvus temperature (Figure 5a). This phenomenon is accompanied by a decrease in tension and a simultaneous decrease in inter-granular borders energy. This is an enhancing factor for δ → δ + γ transition. Due to the presence of alloying elements, there is an increase in Cr, Si, Mo, W, P concentration in A2-type ferrite lattice and Ni, N, Cu, Mn, C in A1-type austenite lattice. The transition appears to be diffusion-limited [25]. As a consequence, the newly formed austenite assumes a lamellar (island) structure. Out of homogeneous δ-ferrite, a two-phase structure δ + γ is formed, with its components varying from one another in terms of corrosion resistance. Chrome and molybdenum (ferrite formers) strongly increase the electrochemical potential. Consequently, the corrosion resistance is concentrated in ferrite. In austenite, only an interstitial solution of nitrogen significantly can increase the electrochemical potential of steel (Table 1; Figure 5b–d).

**Figure 5.** Metallurgical transformations in duplex steels: (**a**) phase equilibrium diagram and austenitic transformation δ → δ + γ (according to [27]), (**b**) distribution coefficient of alloy additions Kδ/<sup>γ</sup> as a function of temperature (according to [28]), (**c**) typical values of the partition coefficient of alloying elements Kδ/<sup>γ</sup> for supersaturated and water-cooled steels (according to [28]), (**d**) influence of alloying elements on the size of electrochemical austenite potential in stainless steels 18/8 (according to [29]).


**Table 1.** Maximum solubility of alloy additives in ferrite and austenite (according to [26]).

The influence of typical alloying elements on duplex steel corrosion resistance is presented in Table 2. In general, Mn and Ni reduce the corrosion resistance. However, these elements are required due to the strengthening effect of Mn and the need for Ni to initiate a ferrite decomposition and form a two-phase structure.

The process of δ-ferrite decomposition and formation of the δ + γ two-phase structure takes place at temperatures of 800–1200 ◦C. Due to its diffusive nature, the kinetics of the transformation depends on the cooling rate (Figure 6). A slow cooling causes a formation of γ-austenite in the amount close to thermodynamic equilibrium. The distribution of alloying elements between the matrix components is also close to equilibrium in this situation. The rapid cooling, on the other hand, creates a metastable structure with a lower austenite content. It increases the risk of harmful secondary phases precipitation from ferrite supersaturated with alloying additives. The formation of harmful precipitates only within ferrite is a consequence of several dozen times higher diffusion coefficients and many times lower solubility of interstitial elements (C, N) in ferrite compared to austenite [25].

**Figure 6.** Influence of cooling rate on the content of δ-ferrite in duplex steels (according to [30]). Left-hand frame: fast cooling, high δ-ferrite content; right-hand frame: slow cooling, high content of γ-austenite precipitates.

To obtain a higher amount of γ-austenite in the final steel microstructure, it is desirable to cool it slowly in the temperature range of the two-phase structure. It can be ensured by a sufficiently high linear energy of welding. At a temperature below 1050 ◦C, the situation is reversed, and a higher cooling rate is necessary to counterbalance the high tendency to secondary phase precipitations so that the cooling line does not cross the upper Time-Temperature-Transformation (TTT) curve (Figure 7).



