Duplex stainless steel. Part 1

Job Knowledge

The name 'duplex' for this family of stainless steels derives from the microstructure of the alloys which comprises approximately 50/50 mixture of austenite and delta-ferrite. They are designed to provide better corrosion resistance, particularly chloride stress corrosion and chloride pitting corrosion, and higher strength than standard austenitic stainless steels such as Type 304 or 316. The main differences in composition, when compared with an austenitic stainless steel is that the duplex steels have a higher chromium content, 20 - 28%; higher molybdenum, up to 5%; lower nickel, up to 9% and 0.05 - 0.5% nitrogen. Both the low nickel content and the high strength (enabling thinner sections to be used) give significant cost benefits. They are therefore used extensively in the offshore oil and gas industry for pipework systems, manifolds, risers, etcand in the petrochemical industry in the form of pipelines and pressure vessels.

In addition to the improved corrosion resistance compared with the 300 series stainless steels duplex steels also have higher strength. For example, a Type 304 stainless steel has a 0.2% proof strength in the region of 280N/mm2, a 22%Cr duplex stainless steel a minimum 0.2% proof strength of some 450N/mm2 and a superduplex grade a minimum of 550N/mm2.

Although duplex stainless steels are highly corrosion and oxidation resistant they cannot be used at elevated temperatures. This is due to the formation of brittle phases in the ferrite at relatively low temperatures, see below, these phases having a catastrophic effect on the toughness of the steels. The ASME pressure vessel codes therefore restrict the service temperature of all grades to below 315°C, other codes specify even lower service temperatures, perhaps as low as 250°C for superduplex steels.

Duplex alloys can be divided into three main groups; lean duplex, 22%Cr duplex and 25%Cr superduplex, and even higher alloyed, hyperduplex grades have been developed, this division being based primarily on the alloy's alloying level, eg in terms of 'PREN' (pitting resistance equivalence number), a measure of the alloy's resistance to pitting corrosion. PREN is calculated from a simple formula: PREN = %Cr + 3.3%Mo +16%N and an allowance for W is sometimes made, having a factor of 1.65. A duplex steel has a PREN less than 40; a superduplex a PREN between 40 and 45 and hyperduplex a PREN above 45, whilst the lean grades typically have lower nickel and hence lower price.

The commonest shorthand method of identifying the individual alloys is by the use of the trade name, particularly for the superduplex grades, eg UR52N+, Zeron 100, 2507 or DP3W, whilst the most common 22%Cr grade, UNS S31803 has widely become known as 2205 regardless of its supplier, although this is a trade name.

The UNS numbering system offers an independent alternative. Typical compositions and minimum proof strengths of the more common duplex alloys are given in the Table. Note that the commonly used 2205 applies to two UNS numbers, S31803 and S32205, with S32205 being a more recent and controlled composition.

Typical compositions and proof strengths of common duplex stainless steels

CommonName

UNS NoBS EN

NoSteelType

Typical Chemical Composition % 0.2%proof

strengthN/mm2 (min)

%C Cr Ni Mo N Cu

2304 S32304 1.4362 duplex 0.015 23.0 4.0 0.055 0.13   4002205 S31803 1.4462 duplex 0.015 22.0 5.5 3.0 0.14 - 4502205 S32205 1.4462 duplex 0.015 22.5 5.5 3.3 0.17   450255(UR52N) S32520 1.4507 super duplex 0.015 25.0 7.0 3-5 0.28 0.13 5502507 S32750 1.4410 super duplex 0.015 25.0 7.0 4.5 0.28 0.3 550Zeron 100 S32760 1.4501 super duplex 0.015 25.0 7.0 3.5 0.25 0.8 550Sandvik SAF3207

S33207 - hyper duplex 0.03 31 7.5 4.0 0.50 0.75 700

The metallurgy of the duplex stainless steel family is complex and requires very close control of composition and heat treatment regimes if mechanical properties and/or corrosion resistance are not to be adversely affected. To produce the optimum mechanical properties and corrosion resistance the microstructure or phase balance of both the parent and weld metal should be around 50% ferrite and 50% austenite. This precise value is impossible to achieve repeatably but a range of phase balance is acceptable. The phase balance of parent metals generally ranges from 35 - 60% ferrite.

Whilst composition and, perhaps more importantly, heat treatment parameters are relatively easy to control this is not the case during welding. The amount of ferrite is dependant not only on composition but also on the cooling rate; fast cooling rates retain more of the ferrite that forms at elevated temperature. Therefore to minimise the risk of producing very high ferrite levels in the weld metal it is necessary to ensure that there is a minimum heat input and therefore a maximum cooling rate. A rule of thumb is that heat input for duplex and superduplex steels should be not less than 0.5kJ/mm although thick sections will need this lower limit to be increased.

Welding consumables are also generally formulated to contain more nickel than the parent metal, nickel being one of the elements that promotes the formation of austenite. A duplex filler metal may contain up to 7% nickel, a superduplex up to 10% nickel.

Reference to the phase diagrams and CCT curves shows that the duplex stainless steels fall within the area where the production of brittle intermetallic phases is a major risk during welding and heat treatment, markedly reducing both toughness and corrosion resistance.

The main culprits are sigma phase, chi phase and 475°C embrittlement. Sigma and chi phases form at temperatures between 550 and 1000°C with the fastest rate of formation around 850°C. The time to form these phases can be as short as 30 or 40 seconds in a superduplex alloy. 4750C embrittlement, as the name suggests, occurs at lower temperatures of some 350 - 550°C with times for the start of formation of perhaps 7 - 10 minutes.

Short times such as these are within the ranges that may be encountered during interpass cooling so, once again, heat input and cooling rates become very important welding parameters except that this time it is the maximum heat input that needs to be controlled. A maximum heat input of 2.5kJ/mm should be acceptable for the duplex steels and 2.0kJ/mm maximum for superduplex. Many codes and contract specifications, however, further restrict heat inputs to less than 1.75 - 2kJ/mm for duplex steels and 1.5 - 1.75kJ/mm for superduplex.

Two other factors that also affect cooling rates are preheating and interpass temperatures. Preheat is not generally regarded as necessary for duplex stainless steels unless the ambient conditions mean that the steel is below 5°C or there is condensation on the surface. In these situations a preheat of around 50 - 75°C should be adequate. Very thick section joints, particularly those welded with the submerged arc process, can also benefit from a low preheat of around 100°C.

Interpass temperature can have a significant effect on the microstructure of the weld and its heat affected zones. For a duplex steel 250°C is regarded as an acceptable maximum and for a superduplex 150°C maximum. Note, however,

that many codes do not separate the grades into duplex and superduplex and 150°C is often required as the norm. Such low interpass temperatures can have a serious effect on joint completion times and forced cooling by blowing dry air through the bore of a pipe once the bore purge has been removed has been used. This is generally only beneficial when thick wall vessels or pipes are being welded using a rotated pipe mechanised TIG process or submerged arc. If this technique is used then it is advisable to force cool the procedure qualification test piece to ensure that cooling rates (and the resultant microstructures) are within the permissible range.

Care therefore needs to be taken to read through code and contract specification requirements and to ensure that the requirements with respect to heat input, interpass temperature etc. are incorporated in welding procedure documentation prior to welding procedure qualification. The next Job Knowledge will provide some guidelines for the welding of the duplex stainless steels.

Duplex stainless steel - Part 2

Job Knowledge

Part 1

The previous article highlighted some of the problems encountered when welding duplex and superduplex stainless steels, in particular the need to control closely the heat input if an undesirable phase balance or the formation of brittle intermetallic phases are to be avoided.

This requirement has implications with respect to quality control. Variations in weld preparations which would be compensated for by the welder changing his welding technique, wide root gaps for example, may result in a significant change in heat input. Weld preparations therefore need to be more closely controlled than for a conventional stainless steel.

It is recommended that weld preparations are machined for greatest accuracy but, if hand-ground, close attention must be paid to the weld preparation dimensions. Welding supervisors and inspectors also need to understand the importance of heat input control, ensuring that welding is not allowed to take place outside the limits of the qualified procedures with regular checking of welding parameters and interpass temperature.

Hot cracking is rarely a problem due to the high ferrite content but has been observed, particularly in submerged arc welds. Cleanliness of the joint is therefore still important. Machining or grinding burrs and any paint should be removed and the joint thoroughly degreased and dried prior to welding. Failure to do so can affect corrosion resistance and joint integrity.

Hydrogen cold cracking, whilst unusual, is not unknown and can occur in the ferrite of weld metal and HAZs at quite low hydrogen concentrations. It is recommended that the hydrogen control measures used for low alloy steel consumables should apply for duplex consumables. Submerged arc fluxes and basic coated electrodes should be baked and used in accordance with the manufacturer's recommendations; shield gases must be dry and free of contaminants.

Most commercially available welding consumables will provide weld metal with yield and ultimate tensile strengths exceeding those of the parent metal but there is often difficulty in matching the notch toughness (Charpy V) values of the wrought and solution treated base metal.

TIG welding gives very clean weld metal with good strength and toughness. Mechanisation has substantially increased the efficiency of the process such that it has been used in applications such as cross-country pipelining.

Gas shielding is generally pure argon although argon/helium mixtures have given some improvements by permitting faster travel speeds. Nitrogen, a strong austenite former, is an important alloying element, particularly in the super/hyper duplex steels and around 1 to 2% nitrogen is sometimes added to the shield gas to compensate for any loss of nitrogen from the weld pool. Nitrogen additions will, however, increase the speed of erosion of the tungsten electrode. Purging the back face of a joint is essential when depositing a TIG root pass. For at least the first couple

of fill passes pure argon is generally used although small amounts of nitrogen may be added and pure nitrogen has occasionally been used.

TIG welding may be performed without any filler metal being added but is not recommended on duplex steels as the corrosion resistance will be seriously impaired. Filler metals are be selected to match the composition of the parent metal but with an additional 2 to 4% nickel to ensure that sufficient austenite is formed. Any stray arc strikes will be autogenous and must be removed by grinding.

MMA welding is carried out with matching composition electrodes overalloyed with nickel and either rutile or basic flux coatings. Basic electrodes give better notch toughness values. Electrodes of up to 5mm diameter are available with the smaller diameters providing the best control when welding positionally.

MAG welding is generally carried out using wires of 0.8 to 1.2mm diameter, rarely exceeding 1.6mm and of a similar composition to the TIG wires. Shielding gases are based on high purity argon with additions of carbon dioxide or oxygen, helium and perhaps nitrogen. Because of the presence of carbon dioxide or oxygen the weld metal notch toughness (Charpy V values) are less than can be achieved using TIG. Microprocessor-controlled pulsed welding gives the best combination of mechanical properties. Mechanisation of the process is easy and can give significant productivity improvements although joint completion times may not be as short as anticipated due to the need to control interpass temperatures to below the recommended maximum.

Flux-cored arc welding (FCAW) is used extensively with major productivity gains being possible in both manual and mechanised applications. The flux core is generally rutile; the shielding gas CO2, argon/20%CO2 or argon/2%O2. The presence of carbon dioxide or oxygen leads to oxygen, and, in the case of CO2, carbon pickup in the weld metal, thus notch toughness is reduced. Metal cored wires are also available that require no slag removal; better suited to mechanised applications than flux-cored wires. Because of differences in flux formulation and wire composition between manufacturers it is recommended that procedure qualification is carried out using the specific make of wire used in production even though the wires may fall within the same specification classification.

Submerged arc welding (SAW) is generally confined to welding thick wall pipes and pressure vessels. Solid wires, similar to those available for TIG welding, are available. Fluxes are generally acid-rutile or basic, the latter giving the best toughness values in the weld metal. As with any continuous mechanised welding process the interpass temperature can rapidly increase and care needs to be taken to control both interpass temperature and process heat input. Because of the need to control heat input the wire diameter is normally limited to 3.2mm permitting a maximum welding current of 500A at 32V although larger diameter wires are available. However, any productivity gains from the use of a large diameter wire and high welding current may not be realised due to the need for interpass cooling.

There is often the need to weld duplex/superduplex steel to lower alloyed ferritic steel, a 300 series stainless steel or a dissimilar grade of duplex steel. The 300 series stainless steels are generally welded to duplex steels with a 309MoL (23Cr/13Ni/2.5Mo) filler metal. Low carbon and low alloy steels may be welded to duplex steels using either a 309L (23Cr/13Ni) or a 309MoL filler metal.

These two filler metals, however, have yield and ultimate tensile strengths substantially less than most low carbon/low alloy steels and all duplex steels. This means the designer has to take this reduction of strength into

account by increasing the component thickness or the welding engineer has to select a filler metal that both matches the strength of the weaker steel and is compatible with the two parent metals. These considerations narrow the choice to one of the nickel-based alloys such as alloy 82 or, for higher strength, a niobium-free high alloyed nickel filler, such as C22. or 59. Alloy 625 has been used but problems with reduced toughness due to the formation of niobium nitride precipitates along the fusion boundary have resulted in the alloy falling out of favour.

Duplex steel welds are seldom post-weld heat treated. Due to sigma phase formation they cannot be given a heat treatment at the low temperatures of 600-700°C, the normal range for stress relief unless a qualification programme has been undertaken to demonstrate that the loss of toughness is acceptable. If PWHT is required then ideally the whole component must be given a solution anneal at 1000-1100°C followed by a water quench; an impractical operation with most welded structures.

Lastly, any process that heats the steels above 300°C will affect the mechanical properties. Heat straightening to control distortion should therefore not be carried out. The HAZs produced by hot cutting processes like plasma or laser may contain undesirable microstructures. Cut edges that will enter service 'as-cut' must be ground or machined back for a minimum of 2mm to remove the HAZ and ensure there is no loss of toughness or corrosion resistance.

If the cut edges are welded after cutting then the HAZs are generally sufficiently narrow that the effects of the cutting operation are lost although it is recommended that, as above, the edges are ground or machined back 2mm.