2507 superduplex welding residual stress relieving and permanent deformations

[MGZmechanical]

Hi, we have made a valve for water at low pressure (2 bar) in superduplex 2507 welding thick parts (100 mm - 4"). The valve stress levels are very low (100 MPa at test pressure 3 bar). But we are getting permanent deformations and each time we make a test at 3 bar it deforms a bit more affecting the sealing. The only thing I can think as root cause is that there are residual stresses in the weldings and each time we pressurize the system it relaxes them. I don't know if someone have some kind of experience with similar problems. Is it reasonable my supposition? Which methods do you use to relax residual stresses in superduplex? Vibration? Cycles of hydro test? Hydro test at much higher pressure? I see PWHT are forbidden in all the literature I've seen.

The valve has been welded with a proper PQR-WPS-WPQ and the base materials and fillers have all the certificates checked.

Thanks in advance.

 

[gtaw]

I don't know that PWHT is forbidden, it is just that the heat treatment is more involved than bringing it up to temperature and then allowing it to cool as one might do with carbon steels.

Duplex stainless is heat treated to begin with. It is heated to eliminate Sigma phase and quenched to produce both ferritic and austenitic microstructures in the correct proportions. I would assume a similar heat treatment would be followed to mitigate the formation of Sigma phase that is likely to form if held at a temperature as if one were stress relieving carbon steels. This is where a metallurgist comes into the picture. That is their business, how to properly heat treat various metals.

Here's a link you might find useful.

https://www.theworldmaterial.com/duplex-2205-stainless-steel-uns-s32205-s31803-alloy-material/

Best regards - Al

 

[CWEng]

In theory, a pressure test may be able to produce slight deformations, as residual welding stresses when present can be up to the yield strength of the material. When you hydro test, if the residual stresses are within 100MPa of the yield strength, you could get yielding in those locations where residual stresses plus hydro test stresses combine to put the stress over the yield point.

However, if the valve has already been welded, the welding procedure would not likely be qualified for a PWHT. In my past life, we welded a lot of duplex sst valve castings, and we would solution anneal repairs a lot of the time, but a lower nickel content in the filler metal is normally required to maintain the ferrite-austenite balance in the filler metal after solution annealing.

Are the welds you are talking about repair welds, fabrication welds, or some sort of overlay?

 

[MGZmechanical]

Hi, the welds are fabrication welds.
All the information I have read till now discourage heat treatment of superduplex stainless (that's why I used forbidden maybe too soon).

Just one note that maybe was not clear in my first message. The stress of 100 MPa is the calculated stress obtained by FEM. We have not measures actual stresses.

 

Don't even think about stress relieving superduplex or duplex, you will destroy it. Service temperature for 2205 is limited to 600°F.

For the deformation after hydro test problem, I would first rule out all mechanical factors. You have a shape and/or integrity issue, the metallurgy of the weld has nothing to do with it.

"Everyone is entitled to their own opinions, but they are not entitled to their own facts."

 

[EdStainless]

The only HT option is solution anneal, and the requirements for temperature uniformity and aggressiveness of the quench are both very strict (if you want good results).
I am concerned if you are seeing distortion in Hydro, you should be way below yield. Unless there are some dimensional or geometry issues I don't see how this is happening.
I would give it a few cycles and see if it stops, if not you need to look at the mechanical side of this.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[mfgenggear]

a quick google of a manufacture of the material data sheet , the mfg recommends post anneal after welding. click on the link.
but has a confusing note:

Heat Treatment Solution Treatment (Annealing) Heat to 1040 - 1120°C and cool rapidly in air or by water quenching. Duplex and super duplex grades cannot be hardened by thermal treatment.
Welding 2507 is weldable by standard methods, without pre-heat. Consumables over-alloyed with nitrogen and nickel are generally recommended such as those with ISO designation “25 9 4 L N”. TIG (GTAW), MIG (GMAW) and all positional manual (MMAW) electrodes are available. Heat input should be within the range 0.5 – 1.5kJ/mm. Post weld annealing is essential following autogenous welding, but not otherwise. Nickel-based consumables (eg Alloy C22) can also be used to give higher corrosion resisting welds. As for other duplex stainless steels the coefficient of thermal expansion of 2507 is lower than for austenitic grades, reducing distortion and residual stresses.

link:

http://www.atlassteels.com.au/documents/Atlas Grade datasheet 2507 rev May 2008.pdf

 

[EdStainless]

PW anneal should always be used with autogenous welds, and it is an option with welds with filler.


= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[mfgenggear]

OK I get thanks Ed, had to educate my self , but is an option for homogenous & heterogenous if the above recommended temperatures.
in the note above but Quote "[highlight #EF2929]not other wise[/highlight]" crazy

I would be skeptical to quench in water , bad mojo happens with water as a quench medium.

 

[0707]

"Pre-heating
Preheating of duplex steels is not recommended and can have a
detrimental impact on the properties of the welded assembly. To
remove moisture, the duplex steel can be heated up to 100°C
(210°F).
Interpass temperature
As a general rule, the interpass temperature should be limited to
150°C (300°F) for lean duplex and duplex steels, and 100°C
(210°F) for super-duplex steels. Too high interpass temperatures
will decrease the cooling rate. In this case, intermetallic phase,
nitrides or carbides precipitation is likely to occur in the weld metal.
Post-heating
Post-heating is not recommended and must be avoided if the
temperature exceeds 290°C (550°F)"

https://industeel.arcelormittal.com...teels-Welding-Guidelines-EN-Juin-2019-Web.pdf

lm

 

If welding duplex SS were as simple as following interweb tips I would lose a lot of business.
Of all the alloy types that should never be attempted by amateur WEs, duplex is the first.

"Everyone is entitled to their own opinions, but they are not entitled to their own facts."

 

[mfgenggear]

@MGZMechanical
it seems more inspection is required.
option: temporary serialize & dimensionally inspect & record before hydro test
then after Hydro Test reinspect and record dimensionally.
reason: verify if there is a dimensional change . to get reliable data this is what I do
for any critical process that causes distortion as dimensional change and have record of it.

if there is no dimensional change then there is elastic deformation, if there is then there is plastic deformation.


create a fishbone chart as the result if there is dimensional change
a key is to prevent residual stresses which can cause dimensional change
some tips to look excessive feeds and speeds in machining, it is human nature for cost
reduction to maximize heavy feeds and speeds, reduce stresses by reducing feed & speeds.

during machining & welding is there fixed process so procedures can not be changed.
it can slightly change from even from one fabricator to the next.

sourcing or first article on site inspect of the first part should be done, and recorded.
thus it has been tool proofed. start an investigation for each step of the process.
it might not even be weld related.

 

[EdStainless]

I have seen 6-12" thick SDSS forgings quenched in highly agitated iced brine trying to get them to cool fast enough to suppress secondary phase formation.
The residual stresses are another issue ...

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[MGZmechanical]

Thanks everyone for your help. We will make some more tests to check it

 

[mfgenggear]

Let's know how it was resolved

 

[MGZmechanical]

For the moment we have stopped the tests. Will inform on september whe we cotinue with the tests