CD4MCU welding

[mahadhatu]

Dear Gurus

I am trying to set a PQR for CD4MCU duplex stainless steel casting. After welding, the ASTM standard was followed for PHWT which is, stabilising at 2050F for 1in/hr and 2h min and then furnace cooling to 1900F, stabilising at 1900F and quenching in agitated water. I came up with an average hardness of 237BHN. When subjected to a Guided Bend Test the material failed terribly and fractured without undergoing any bending.

I would really appreciate if some one could throw some light on where I am going wrong.

Thank you

 

[GRoberts]

First, if you could provide some more relavant information, it would be helpfull.
1. How thick are the test plates/pipes you are welding?
2. How was the material heat treated before welding (times and temperatures)?
3. What welding process and filler metal are you using?
4. Where did the bends fail? Base/HAZ/weld?
5. What sort of results and/or failure mode did you get on the tensile tests?
6. What were the actual times and temperatures for the PWHT including the 1900F? What equipment are you using for quenching?

If you know where the bends failed, the first thing I would do to gain some valuable information is to polish and etch a weld cross section and look at the area in question for intermetallic phases or other problems.

 

[mahadhatu]

Thank you GRoberts

1. Thickness of plates = 1 inch
2. Heat treatment before welding :
Cycle 1 : Raise temperature at 200F/hr up to 2150F
Hold for 1hr/in
Furnace cool to 1900F
Hold for 1hr
Air Cool
Cycle 2 : Raise temperature to 550F stabilise
Raise temperature at 400F/hr up to 1900F
Hold 1 hr
Water Quench within 3 mins
3 SMAW with Ferralium 255 electrode was used without preheating and interpass temperature was 350F max
4. The welds failed in the weld bead area in a Guided Bend Test. The fracture shows a dark gray surface with Bright gray spots scattered all over.
5. The testing was stopped right after it failed bend test
6. PHWT: Raise temperaure to 2050F
hold 1hr/in, 2hr min
lower to 1900F in FCE
Hold 1 hr
water Quench
Water quenching was performed in a water tank agitating with air bubbles.

As you rightly mentioned, there might be intermetallic phases being formed in the weld which is evident from the (2) colored fractured surface.

We performed the guided bend test on a strip of just the base metal and the results were satisfactory with no sight of cracks.

 

[GRoberts]

Well your answers have eliminated several possiblities, so the focus needs to be narrowed down to the filler metal and it's heat treatment.

If possible, get a chemical analysis of your deposited filler metal just to make sure nothing is awry there. Some additional helpful information would be to perform a ferrite check on the weld. If it is significantly different than predicted by the WRC diagram, it would be further indication of intermetallic phases, usually sigma. Although I have more experience (i.e. welding procedures that failed the first time) with 2507 type duplex (ASTM A890 Gr. 5A), the same kind of principles apply. If you are going to try the test again, a couple of suggestions are to hold the 2050F a little longer. If there was significant sigma that formed during heating, or during welding, it may take longer than 2 hours to dissolve. 4 hours should be enough for a thin plate. I think a more likely culprit is the 1900F step cooling. I have done welding tests where a weld cooled from 2075F and held at about 1950F for 8 hours wouldn't pass bend tests(and had bad charpy values), but one held for just 2 hours bent fine. This was due to intermetallics. I have also seen it reported in literature that some high moly stainless steels can have intermetallic phases form at temperatures above 1900F.

Since the 1900F step cool is designed to keep the yield strength up, (Quenching from 2050F would result in low yield), see if you can raise it to 1925F or 1950F and live with the possiblity of a slight yield strength reduction. If my memory serves me, ASTM has 1900F as a minimum only, no max? Also try and keep it as short as possible based on the needs of the casting you will be welding.

With a 1" thick plate, quenching is probably not a main issue, but if you can get a better quech by agitating with propellers, or making sure the water doesn't get too hot, that may help also.

 

[bruv]

I have a couple of comments:

I am surprised that there is no sigma in the parent plate. I'm guessing, but I think that an air cool from 1900F would form sigma. The fairly fast re-heat (400F/hr), followed by a short hold may not re-dissolve the sigma. Holding for up to 3 minutes will also not help. After PWHT, what is your time from oven to quench tank?

GRoberts ideas of increasing the time at 2150F and reducing the hold time at 1900F are good. You may be able to reduce it to as little as 15 to 30 minutes.

Good luck

 

[mahadhatu]

GRoberts, bruv,
Thank you for your comments.

Following your suggestions, I redid the heat treatment, this time heating to 2100F, holding for 4 hours, cooling in furnace to 1925F, holding for 1 hour and quenching in mechanically agitated water in less than (1) minute.

The test results showed that the hardness in was still on the higher side at 262BHN in weld and base metal (desired - 255BHN)

The weld failed one of the 2 bend tests with cracks larger than 1/8". this time the failure was not catastrophic.

Yeild(80Ksi), Tensile(104 Ksi) and Impact (weld:32, HAZ:48, Base:60 ft-lb) properties are well above desired.

So the question remains as to what else can be done to bring the average hardness to lower than 255BHN?
Since it still failed the bend test, there are still traces of intermetallics remaining after this heat treatment.
I suspect that holding at the 1925F for a longer time might help. Any comments?

Thank you

 

[GRoberts]

What is the chemistry of the weld? It will be softer and more ductile if you have more austenite. It is difficult for manufacturers of SMAW electrodes meeting the AWS A5.4 specification for 2553 electrodes to keep the ferrite down (or austenite up). For some reason there is a lot lower maximum for Ni in the SMAW specification than the FCAW specification. Have you done any ferrite checks on the weld? Also, did you do any sort of NDT beforehand? Could the failed bend test be at a welding defect?

Holding at 1925F longer won't help unless the whole plate didn't reach the temperature, but it should have, being only 1" thick.

To check for remaining intermetallics, polish and etch a cross section of the weld to look for them.