notch toughness failure

[saipemsmspetti]

In this moment I am working as welding engineer on Sakhalin pipeline project. In the last few weeks we are faced with constant notch toughness failures in welding qualification program. In spite of all my experience there is lack of time and possibility to explore the problem in details and pressure to continue with on site works.

The short description of the problem:

Base material: API 5L X65 pipe 48"x34.4 mm
(CE=0.34%, PCM=0.15%)
GTAW (root)+ SMAW (5G position, uphill)
ER70S-6 + E8018-G
Heat input:Q=15-18 KJ/cm (for root and 1&2 fill)
Q=22-38 KJ/cm (to complete)
Preheat: 125 deg. C (Used induction heaters to preheat whole coupon and to keep preheating temperature all the time)
Interpass: 250 deg. C
PWHT: At 580 +/- 20 deg C for 1.5 hour soaking

We have constantly repeated CVN failures but only in the root weld metal and in the root fusion line (10x10 CVN specimen taken 2 mm in depth from the root surface).
Testing temperature: -35 deg C (CVN permitted av.=45 J, min=34 J)
In the same time hardness values are good across the whole welding joint (260 HV) as a result of PWHT.
To investigate a problem we ordered microstructure photos of this part of the weld joint (as minimum waste of time) to make some decisions.
Waiting the respond of this test, I am afraid of what should I be able to decide and what to expect.
Can I expect a grain coarsening as the main reason, or some precipitates on the grain boundaries or perhaps some martensitic or bainitic structure?
Is the problem with to much time on a peak temperature of welding cycle or too small cooling rate between 800 and 500 deg C?
Is PWHT that one to deteriorate toughness, but only in the root region.
What should be the right step to improve toughness and solve this problem


Thank you very much in advance

 

[metengr]

saipemsmspetti;
What is the actual chemical composition of the SFA 5.5 E8018-G weld rod? Why are you even using this weld rod, when no impacts are specified?

Also are you aware that for ER70S-6 electrode and wire supplied in accordance with SFA 5.18, minimum impact requirements as-welded is 27J at -30 deg C.

 

[saipemsmspetti]

Thanks for urgent response.

E8018-G is BOHLER electrode FOX EV 65 with:
C=0.06
Si=0.3
Mn=1.2
Ni=0.8
Mo=0.35
with 80J at -60 deg. C from BOHLER catalog

ER70S-6 is ESAB rod OK Tigrod 12.64 with:
C=0.1
Si=1.0
Mn=1.7
with 47J at -40 deg. C or
with 70J at -30 deg. C per ESAB catalog

If I am concerned about consumable choice it is maybe OK Tigrod 12.64. I thought OK Tigrod 13.13 would be better choice. This is one of my ideas to try to solve this problem. But I think FOX EV 65 is good choice because the results of the rest of weld section were with very good notch toughness.

 

[unclesyd]

I saw a program on the Discovery Channel about the LNG project and one of the scenes made me wonder about the pipe welding.
The scene was of a TIG welder making what looked like a third pass, root + hot + 1 additional, using the "walking the cup technique". This in itself usually falls under it's OK but I have seen some problems on welding using this technique where the welder dwells too long. This is especially true on wide welds.
It has been our policy to limit the "walking the cup" technique to the root and hot pass only on alloy steel piping.

 

[saipemsmspetti]

Thanks 'unclesyd' for your contribution to solve my problem.
I do not know about this application of TIG as a third pass. What should be the purpose of it? Usualy TIG is used for the root and after from the hot to the cup passes SMAW is used.
But you gave me an idea with "walking the cup technique".
I suspect in grain coarsening present in the root weld metal and HAZ. Maybe is this technique with dwelling too long by sides of root pass that caused grain growth. I will see the microphotos of this region that I am expecting to get soon to check it. Your last sentence said that for X65 we should have be very careful with using TIG with such technique.

 

[metengr]

I would request a metallurgical examination of the weld root CVN impact surfaces as well. It might be that the lower impact energy values of the weld metal deposit are attributed to weld defects (I have seen this before). You need to verify this to eliminate it as a possibility along with checking/verifying the weld rod/electrode compositions. I have seen mislabeled weld rod cans and other stuff from suppliers.

 

[stanweld]

What is the chemistry of the X65 pipe? Dilution of the root pass may be a partial cause of your problem. What is the travel speed? The only cases that I have ever had where low impacts were obtained in the root with GTAW involved travel speeds < 2.5 in/min.

 

[GRoberts]

Are you sure that the published values are with PWHT? Previosly I have used ER80S-D2 for GTAW root with sucess after PWHT with CVNs. A test report I have from ESAB (Spoolarc 83) reports 199 ft-lb average at -20F after PWHT of 1125F for 24 hours, and a -80F NDTT temperature. I think with the lower PWHT temperature, your results will be lower, as the weld deposit will not recieve as much tempering. My results on pipe with 1000F PWHT for 3.5 hours was 79 ft-lb at 32F.

I would also agree with Stanweld to check your dilution. The root pass composition could be quite different than the manufacturer's analysis. Make sure plenty of filler wire is added to the root pass to help with this potential issue.

 

[unclesyd]

A couple of other points about "walking the cup" welding technique:

A good welder will want to use a "feather edge" weld prep, no gap, for the root. We haven't seen any problems with this approach and use it for all piping in polymer service. In fact this technique is used in 90% of piping on site, again only the root + hot with SMAW fill and cap.

Welders that are proficient with the process will want to completely weldout the joint if possible. The don't like to "stick" weld.

Some welders can and will carry a tremendous amount of weld metal with this process. They will double by twisting the stated wire size.

I agree the a quick macro etch might tell a lot about the microstructure. I would also get some micro samples.

 

[saipemsmspetti]

Thank you 'unclesyd' for your point of view and contribution, but let me ask you some details to make me clear your respond.
We have 3.5-4.2mm root gap which means less dilution in the root then with no gap. Do you prefer as much as possible less dilution and smaller amount of weld metal in the root.
Please can you explain me in more details your sentences and terms in quotations because it is some sort of sleng and my english is not native.
-"feather edge"
-"walking the cup"
-weldout the joint
-"stick" weld.
-double by twisting the stated wire size.

 

[saipemsmspetti]

Thanks 'GRoberts'.
I think I will follow your instruction to use ER80S-D2 or any other with properties after PWHT.
Your words 'Make sure plenty of filler wire is added to the root pass to help with this potential issue' mean that it is necessary to have such technique to diminish dilution with as much as possible shear of filler metal in the root pass and less melting of BM.
This is somahow in contradiction with 'unclesyd's opinion.
Why do you mention lower PWHT temperature and lower noch toughness results.Should I have to make PWHT on higher temp. than 580 deg. C to get some better results.

 

[saipemsmspetti]

Thanks 'stanweld', you were pointed a new fact for the problem.
It is true. We did had 4.65 - 6.22 cm/min speed for GTAW.
Regarding chem. content of X65 find it below and tell me your opinion:
C=0.077/Si=0.282/Mn=1.61/Mo=0.014/Ni=0.272/Cr=0.041/V=0.075/Cu=0.218/Nb=0.045/Ti=0.002/P=0.014/S=0.00008/N=0.0038 and small amounts of As,Sb,Ca,Bi
Speaking of dilution can you explain which process is taking place in your opinion during root pass, regarding diffusion and changing the chenical content of this region.
Who is the main to blame for dilution.

 

[metengr]

saipemsmspetti;
You have been given some excellent suggestions that may or may not solve your problem. I would wait until you receive the results back from the lab to confirm one or more of the suggestions mentioned above.

 

[saipemsmspetti]

Thanks 'metengr'.I agree.I think I am much closer to solve the problem.
Anyway I will try to use other rod for GTAW (with PWHT garanteed values)and take a care about welders skill and technique.
Please, contribute once more with your opinion, after receiving all asked clearances from the others.
It is very good sentence of yours 'may or may not'.Just like Pitia the Greek prophet woman. (Ibis, redibis nunquam in bello peribis)

 

[metengr]

saipemsmspetti;
"May or may not" is a very popular phrase around my house especially with my two daughters that attend college.

 

[GRoberts]

Typicaly, base metal is higher in carbon than the weld metal. The extra carbon in the weld metal from dilution typically does not help CVN values.

Higher PWHT temperature in low alloy steel (at least up to 1100F minimum) in my experience seems to help CVN values. PWHT below 1100F can be challenging. As an example, with E12018M2, we get poor CVN values with 1075F PWHT, and great values with 1175F PWHT. Other alloys may not be as extreme, but seem to follow the same trend when we test CVN properties. The exception may be carbon steel FCAW, which starts to degrade above 1150F or more. I would temper as high as the base metal allows (i.e. 25F-50F below base metal tempering temperature). If pipe is as-rolled, you may need to conduct some experiments to see how it responds to different temperatures.

 

[unclesyd]

"Walking the cup" is welding technique where the welder lets the shielding cup of the torch actually touch the metal and walks it back and force. Really proficient welders can weld in similar manner without the shielding cup touching. In the proper hand it is a very efficient method on putting in a very good and smooth root pass. On the hot pass he can control the heat to prevent burn through. The process requires very good dexterity of the hand and you will see the welder rapidly working his torch. This method was first used on SS and higher alloy piping and has slowly migrated to CS and low alloy piping.

"Feather edge" This is weld prep where the their is no land on the weld prep, it is one single angle. The edges are touching and there is no gap. The welder likes the surface of the weld prep to be very smooth.

"Weldout the joint" This is a term used to convey the process of completing or finishing the weld joint after the root and hot pass are completed.

"Stick weld" This welding with a covered electrode, SMAW.

"Double by twisting" This is process where one takes two of the same size wires and twist them together to effectively make a large electrode to carry more metal. A lot welders have a little jig that will make a very tight twist of say two 1/16" wires to make a 1/8" equivalent.


Though my experience with low alloy pipe is some what limited we tended to minimize the dilution in root pass of these alloys. Fortunately we never had any requirements for low temperature high impact values other than Ammonia Piping.


I forgot to mention that you have one heck of a project underway. I know why you you need the low temperature impact values as I was stationed across the Sakhalin straight on Northern Hokkaido for several months during the middle of winter while in the military.

 

[stanweld]

The Nb and V content are somewhat high in the base metal. Dilution could be a problem. PWHT can lead to carbide/carbonitride precipitates with resultant lowering of impact toughness.

 

[grampi1]

Agree with high Nb+V possibily contributing to the problem.

Look for LBZ's Local Brittle Zones along the Fusion Line which would show up as harder areas vs the majority of the HAZ. You have low Pcm but are still getting 260HV with very reasonable arc energy. I would have guessed hardness below 250. The high Nb+V could be responsible through secondary hardening.

Are you sure that preheat is being applied and that interpass is maintained above the required preheat until completion?

Were weld metal consumables batch tested for impacts in the PWHT condition you specified, before purchase? Were consumable brand and designation required supplementary essential variables?

Sometimes it takes a stronger root consumable to match the higher impacts for the higher strength base metal. The typical undermatch in the root I think concentrates strain in the weld metal and makes it less tough than base metal.

 

[stanweld]

Back in the early 1970's major problems with impacts in the root and HAZ were experienced after PWHT when the fine grain melting practice only used V and Nb additions as was apparantly done with your pipe. Al fine grain melting practice with the additions of Nb and V to provide increased strength effectively eliminated the problem. Temper embrittlement is exascerbated in the deposited weld metal with the high Mn in ER70S-6 in combination with the V & Nb through dilution.

I have used ER80S-Ni1 to weld the root in similar pipe subject to PWHT to provide required absorbed energy.