2205 Welding to Shell DEP 1

[InspectorWilly]

All,

I am trying to get some weld procedures for 2205 qualified to Shell DEP 30.10.60.18, and it is not going so well. We have the following parameters:

GTAW Process with 98% Argon/2% Nitrogen. We are keeping heat input below 1.5 kJ/mm with the following variables:

Voltage - 14-15
Amperage - 80-100
Travel Speed - 60-70 mm/min
Interpass - 200F Max

Filler is ER 2209.

We are having problems with ASTM G48 testing on the weld coupons when we use higher heat input (about 2kJ/mm and 250F max interpass)
However, when we used the above variables, corrosion testing was fine per customer spec (<4g loss per m2), but we wind up with lack of fusion in the roots.
Forced Cooling was also used between weld runs to minimize cooling period to below max interpass to ensure minimal formation of delta ferrite.

I am wondering if I use a superduplex filler such as 2594 that has the higher nitrogen content for pitting corrosion resistance and go to pure Argon if that would help my welders get better fusion while maintaining nitrogen content of the weld for pitting corrosion. Any input or experience with 2205 welding with Shell DEP's would be greatly appreciated.

Additionally, everything was about as clean as we could get. Brand new tungsten was used, new gloves, base material and fillers were wiped down with alcohol. Thanks in advance.

 

[pradipgoswami]

Hi Inspector Willy,

The above specification calls for:

The test shall be carried out for 24 hrs immersion at 25±1 °C (77±1 °F) for Duplex steel.
The test acceptance criteria shall be:
- No evidence of pitting examined at a magnification of X 20.
- Weight loss shall not exceed 4g/m2 [~0.04 mm/yr] (1.6 mpy)

As per your query results of corrosion test is fine(weight loss is acceptable as per the spec). However if the issue with lack of fusion, a change of welding technique, especially torch and filler wire manipulation and changes in the groove design could be helpful. But a change of groove design (included angle) may call for adopting the same for production welds, which is sometimes insisted by the clients.

2209 filler is capable of meeting the G48A test criteria as mentioned above, using 2594 filler may be overkilled.It would be advisable to procure the filler metals meeting the above testing criteria from the established manufacturers. 2205(UNS 32205) is a very basic and a less complex duplex S.S.Hence meeting the clients' requirements should not be a major concern.

Use of higher heat input (about 2kJ/mm and 250F max interpass) is not advisable.A typical heat input of 1.2-1.5 KJ/mm and IPT of 200 Deg F are often followed by the vendors in procedure and productioin joints.

Use of Ar+N2(98+2)% as shielding gas looks fine.The backing /purging could be 100% Argon, 100%N2 or Ar+N2(98+2).

G-48A calls for a minimum 0f 120 grit surafce finish for on test specimens. However the corners may require higher grit finish (as discussed with the client).

Hope the above answer clarifies your query. You may see the following attachment for additional guidence.


Thanks






Pradip Goswami,P.Eng.IWE
Welding & Metallurgical Specialist
Ontario, Canada.
ca.linkedin.com/pub/pradip-goswami/5/985/299

 

[pradipgoswami]

Hi Wiilly,

The attachment did not get uploaded.However this is document I was mentioning. Just do a google search:-

WPS America, Welding Practice for the Sandvik Duplex Stainless Steels ,SAF 2304, SAF 2205 and SAF 2507

Presented by

Frank Babish, Sandvik

Thanks.

Pradip Goswami,P.Eng.IWE
Welding & Metallurgical Specialist
Ontario, Canada.
ca.linkedin.com/pub/pradip-goswami/5/985/299

 

[InspectorWilly]

Thanks Pradip,

Your information has been helpful. Since welding 2205 to these requirements (Based in API Technical Bulletin 938-C) is new to us, we are having some issues. I will discuss your recommendations with the welders and see where that gets us.

Thanks

 

[DekDee]

Willy,
Try this

Significance and Measurement of Ferrite in Duplex Stainless Steels

A duplex stainless steel has a structure that is a mixture of ferrite and austenite in roughly equal proportions. It is generally agreed that these phases will have volume fractions in the range of 25-75% for the benefits associated with duplex stainless steel to be achieved.

Ferrite and Austenite

Ferrite, the magnetic structure commonly found in annealed carbon and alloy steels, is moderately strong and has a useful amount of ductility, but it can be lacking in toughness, especially at below ambient temperatures. Ferritic stainless steels are composed of iron, chromium, and sometimes molybdenum. Ferritic stainless steels can have a wide range of corrosion resistance depending on their chromium and molybdenum contents. As a group, they are extremely resistant to chloride stress corrosion cracking.

Austenite, the nonmagnetic structure commonly associated with Type 304 and 316 stainless steels is not as strong as ferrite in the annealed condition, but work hardens very rapidly during deformation, providing very high strength before fracture. Austenite has extraordinary ductility and toughness. Ferritic stainless steels will become progressively austenitic when nickel and sometimes manganese and nitrogen are added. Although austenitic stainless steels have excellent corrosion resistance as a function of their alloy content, they are susceptible to chloride stress corrosion resistance.

When intermediate amounts of the “austenite-stabilizing” elements are added to a ferritic stainless steel, the resulting steel will have a mixture of ferrite and austenite, having excellent strength, good ductility and toughness, and excellent resistance to stress corrosion cracking. This proportionally balanced steel is known as “Duplex” stainless steel. In the annealed condition, the duplex stainless steel mill products are manufactured to have about 40-50% ferrite, with the balance austenite, obtaining the optimal mechanical properties.

Phase Balance in Duplex Stainless Steels

The initial solidification of a duplex stainless steel is completely ferritic, with subsequent partial transformation to austenite regions within the ferritic structure during cooling. It is important to cool the steel slowly enough for the austenite to form in sufficient volume fraction to provide the expected properties. It is also important to cool quickly enough through the range of intermediate temperature, 1300-1800°F, to avoid formation of intermetallic phases, such as sigma. Normal weld thermal cycles are typically too short for any detrimental sigma phase formations to occur. Duplex stainless steel mill products are usually 50-60% austenite, but it is useful to alloy the duplex weld filler materials to obtain a volume fraction of up to 75% austenite to obtain the best toughness while retaining the benefits of a duplex structure.

The use of higher levels of austenite, promoted by nickel, manganese, and nitrogen additions, is typical of all duplex filler metals, but is particularly useful for the flux-shielded welds because the presence of oxygen from the flux tends to reduce the total toughness of the metal. These reduced levels of toughness are acceptable in most applications, but they are not as high as those of typical annealed mill products. The procedures for flux-shielded welds are qualified by demonstrating that they produce adequate toughness. Duplex stainless steels have been welded by a full range of economical welding procedures.

Avesta 2205 and 2507 bare wire (MIG and TIG) is chemically formulated to produce the desired 30-55% ferrite balance when following a qualified welding procedure. The austenite-forming alloy elements (nickel and nitrogen) are balanced with the ferrite-forming alloy elements (chromium and molybdenum) to enable the filler metal to have an undiluted weld metal content of approximately 40% ferrite. But even the duplex weld filler metals that are overalloyed with nickel are capable of producing highly ferritic welds if rapidly quenched, e.g., very small welds on large plates, connecting tubes to tubesheets, small “smoothing” passes, etc.

It is unusual for seam welding to produce excessive ferrite in the modern, higher-nitrogen duplex stainless steels, but it is important to avoid the possibility of excessively fast cooling. These situations are readily avoided or offset by skilled welders. At the other extreme, excessively heat inputs must also be avoided as duplex stainless steels will form intermetallic phases (e.g., “sigma phase”) if the material remains too long in the red heat zone, with 1300-1800°F being the worst range.

Submerged arc welds (SAW) also use bare wire for filler, but SAW should be thought of as requiring the qualified of the combination of filler wire and flux because their interaction determines the properties of the weld. Due to the high heat input incorporated when using SAW, up to 3-5% of chromium, depending upon the heat input, may be lost over the weld arc. Because very little nickel is lost in the weld arc, the chromium loss can lower the ferrite to an undesirable level. Using a chromium-enriched flux can compensate for its loss through the weld arc and allow for more consistent control of the ferrite level.

Importance of Qualified Procedures

The welder must follow a qualified procedure that has been demonstrated to produce good phase balance while avoiding the formation of intermetallic phases in the heat-affected zone. Most often a procedure is qualified by metallography and by demonstrating that corrosion resistance and impact toughness have not been adversely affected. The tests most commonly applied for mill products are those in ASTM A 923. With appropriate adjustment of the acceptance criteria, these tests can also be applied to welds. However, it is essential to recognize that the shelf energy of Charpy V-notch testing for annealed 2205 duplex stainless steel plate may exceed 200 ft-lbs. So a reduction in measured toughness at –40F to less than 40 ft-lb can be used as an indicator of the presence of intermetallic phases in a plate, or in the HAZ. However, the shelf energy of the weld metal, especially those with flux shielding, is commonly 25-30 ft-lb. This lower level, still an indicator of good toughness, does not imply that a faulty welding process was used, as there is no reason to suggest that welds must achieve the same level of toughness as the annealed base metal.

Avesta shielded metal arc (stick) electrodes are formulated with a higher nickel content to produce a lower ferrite content than the bare wire. This higher austenite content will counteract the detrimental effects of the oxygen associated with the flux on the electrode. The higher allowable nickel content for 2209 electrode filler metal is reflected in ANSI/AWS A5.4-92, Specification for Stainless Steel Electrodes for Shielded Metal Arc Welding. At the high end of the permitted range, the nickel content (an austenite-former) will lower the ferrite content to approximately 25-30% ferrite in the stick electrodes. This level will be the minimum achieved in actual welding because of dilution effects and because it is thermodynamically impossible to obtain austenite in excess of its equilibrium level. By following a qualified welding procedure, the desired level of toughness can be achieved with austenite contents that are higher than that of the base metal. This approach associated with the stick electrodes can also be applied to flux core welding because both rely on a flux for the shielding.

Submerged arc welding uses an oxygen-containing flux and so has a tendency to lower toughness. However, the selection of the flux can have important effects on the toughness of the resulting weld. SAW is a high heat input process. It should be applied when the thickness of the section being welded is large enough to provide for two dimensional cooling, with a rule of thumb being that thickness of 5/8-inch or greater is sufficient to assure good use of SAW. Because they are proprietary products, each with their own characteristics, both the filler material and the flux should be specified in the qualification, and the same products should be used in the actual fabrication process.

Other Factors Affecting Phase Balance

Although the ferrite percentage of the filler metal, as established by the composition, is important, there are many other factors having an equal, or greater effect in determining the final phase balance in the weld. To focus only on the ferrite content of the filler metal as the driving force is a mistake. Other factors to consider include excessive nitrogen pick-up (which will significantly lower the ferrite), incorrect cooling rate (as in the qualification weld being made under conditions not typical of practice), dilution with the base material (which may raise the ferrite), improper arc length (improper welding technique), or an extremely low heat input (this can cause rapid cooling and an extremely high ferrite content). Therefore, it is of utmost importance to qualify procedures reflecting actual practices and then to follow those qualified procedures. All of the AvestaPolarit welding consumables have been designed to give adequate toughness in a wide range of “as-welded” conditions provided proper welding techniques are followed.

Measuring Ferrite Content

There are several approaches to determining the “phase balance” of the weldment. The most precise way to determine the ferrite content of a weld metal is a metallographic examination, but this destructive test is time consuming, expensive, and requires a well-equipped laboratory. Percentages of ferrite and austenite in a “matching” 2205 duplex stainless steel weld metal microstructure can be estimated by calculating the chromium equivalents (ferrite formers) and the nickel equivalents (austenite formers) and plotting them on the WRC-92 constitutional diagram. Commercially available portable instruments may be used to measure ferrite in welded joints nondestructively. An example of such equipment is a Magnagage. Because the original instrument only measured ferrite up to 28 FN, a counterweight was incorporated to extend the range to higher ferrite numbers expressed as Extended Ferrite Number (EFN). It is important to recognize the specifications are usually written in terms of volume percent ferrite, but welders frequently use the Ferrite Number, and these are widely different in the range associated with duplex stainless steel. At higher ferrite numbers (above 28 FN), the conversion to volume percent ferrite equals (0.55 EFN+10.6). Therefore, using this calculation, the volume fraction of ferrite for a 70 FN is approximately 40-45%, or nearly ideal for 2205 and 2507 phase balance.

The properties of the weld, including toughness and corrosion resistance, are not readily measured by nondestructive testing methods. The welder must rigorously adhere to the qualified procedures to maintain the desirable properties of duplex stainless steels in the final construction. Deviations from that procedure, even when done with the best of intentions, e.g., the small smoothing pass, can produce a condition of great risk in service. We must remember that following a qualified welding procedure which includes proper weld parameters, heat inputs, interpass temperatures, and cooling rates is essential for the successful welding of 2205 and 2507 Duplex Stainless Steels.


Chuck Meadows
Technical Service Manager
Avesta Welding, LLC

An excellent resource is "Practical Guidelines for the Fabrication of Duplex Stainless Steels" published by the International Molybdenum Organisation - I downloaded it for free years ago and it has helped me immensely.
Cheers,
DD

 

[EdStainless]

Is the qualification test a G48 or A923?
It really should be the latter.
The formation of detrimental intermetallics in the structure is the greatest risk.
These coupons require freshly ground surfaces.
A G48 should be run using the actual service surface conditions. For non-intact surfaces they are usually polished very smooth.
Different test for different reasons.

What is the deposition rate on the root pass? The data looks about right. I would expect it to work.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[jwhit]

A number of specifications are calling out the G48 testing for the duplex materials, NORSOK M630 for instance specifies G48 as do the Shell and Exxon specifications. I'm not certain as to why they choose G48 over A923, just that it is consistently done for these materials. I assume that the people that write the specs know something we don't.

 

[stanweld]

The 2209 filler metal has often been the achilles heel regarding the service life of 2205 in particular environments. Up alloying the filler metal has proved effective in numerous applications.

I only have G48 data with 2594 filler on super duplex that meets Shell criteria.

 

[jwhit]

I have to admit to suffering from a little material confusion. We get orders for A890 Gr.4A castings which sometimes specify corrosion per G48 Method A at 25C, other times A923 and some with no corrosion requirement at all. NORSOK doesn't ask for G48 for the duplex, only superduplex materials. The last Exxon castings were to A923.

 

[NickClark]

Everyone is focusing on G48 but that doesn't appear to be your problem. Although I'm supposed that your passing the G48 with a lack of root fusion as I we could expect this to cause a crevice failure. Can you describe the root gap used and backing gas (mix and Max o2 content). Unless you use a root gap you may struggle with fusion and G48. Its quite common to put a 25 cr root run if you are struggling with G48 but you will then need to keep root run heat input even lower. I suggest you focus on joint design to get fusion without too high a heat input first then look at options to improve corrosion resistance if that becomes an issue.