Residual Stress peaks versus welding technique

[bob330]

Hi All,

Is there any reason to believe that a given carbon steel vessel girth weld made by the SMAW technique with 350F Preheat would be any more susceptible to caustic SCC than the same vessel welded by a SAW technique. I guess another way of re-phrasing this question would be to ask if there any significant difference in the magnitude of internal surface residual stresses generated by the welding. One guy told me that you can expect residual stresses just outside of a girth weldment (and in a direction parallel to it) to be at or near the yield strength of the metal regardless of the welding technique. He did however say that the size of the HAZ and area holding high residual stresses may vary substantially with process/technique but that the development of high residual stress was more or less inevitable. Is this all correct?

Also, since residual tensile strresses are always balanced by compressive stresses, where are the compressive stresses ina girth weld?

Thanks
Bob

 

[vesselguy]

Interesting question.

I am not a welding specialist so I'd leave the answer for SAW vs SMAW to others who know more. However, reqardless of welding process, any vessel in caustic or sour service have to be PWHT for stress relieving to address exactly the SCC concern you have. Hence, since you have to PWHT anyways, what does it matter which process leave less residule stress? Both will come out in the wash after PWHT.

As for developing residule stress, it makes sense that it is a function of how thick your joint is (hence directly related to amount of heat input), and amount of preheating.

 

[bob330]

Thanks VesselGuy,

Actually, due to time constraints, we did not perform a PWHT but did one of the ASME alternate weld techniques with preheat. Was there some risk in this decision......yes indeed, but we took into consideration the fact that the vessel was 40 years old and had never been given a PWHT before and had not experienced any caustic SCC in its history. We will watch it carefully with 18 month inspections.

Bob

 

[metengr]

bob330;
Did you perform the temper bead technique per the NBIC?

Getting back to your OP

I guess another way of re-phrasing this question would be to ask if there any significant difference in the magnitude of internal surface residual stresses generated by the welding.

Yes, there are significant differences in residual stress between welding processes.

 

[robsalv]

Metengr, if it was a multiple pass SAW, would the residual stresses between the SMAW and MAW converge to a similar value???

I don't dabble in much welding - just have a fair-ish working knowledge. I read a paper some time ago that categorically stated that a temperbead process did not stress relieve the weld. The paper said that residual stresses are about the same, but the benefit in a temperbead procedure is that the weld is likely to have significantly lower peak hardnesses.

It's my understanding now, that pre and post heat, and cooling rate, have a much larger bearing on residual stress. I'm also lead to believe that if a temperbead is to address residual stress, then it is required to be temperbead with half grin type process.

I'd appreciate your comments.



In regards to Caustic service, I would be feeling most uncomfortable relying solely on a temperbead process in a serious CSCC environment.

Cheers

Rob



Conventional

 

[metengr]

Rob;

if it was a multiple pass SAW, would the residual stresses between the SMAW and MAW converge to a similar value???

IMO, No.

The temper bead welding was never really intended to provide a sufficient stress relief because for certain benign elevated temperature environments, an in-situ stress relief can ocurr with various grades of steels (those used in power generation). Your statement is correct regarding hardness and temper bead, and it is very effective in reducing peak hardness.

This is why one needs to understand the use of temper bead weld repair techniques and their limitations. If you need to reduce peak residual stresses, you have several choices - vibratory, thermal or using a low stregnth filler metal to reduce weld shrinkage stresses.

IF residual stress is not critical in service and hardness needs to be reduced to avoid hydrogen cracking, the temper bead would perform very well in service.

For carbon steels in caustic service, thermal stress relief and hardness reduction are both necessary.

 

[stanweld]

bob330,
Is the vessel in question in a caustic concentration and temperature range where PWHT is required? If so you will not have relieved stress adequately with the temper bead technique. Your repair weld may well be subject to caustic cracking. The alternative methods defined in the NBIC and API 510 were not fully intended for vessels subject to stress corrosion cracking where PWHT is mandatory.

Reference RD-1010 "Competent technical advice shall be obtained ...such advice being especially necessary if the alternative is to be used in highly stressed areas, if service conditions are conducive to stress corrosion cracking..."

 

[bob330]

Thanks all,

I did want to mention that there is information in the technical literature saying that carbon steel actually becomes more susceptible to caustic SCC as it gets softer believe it or not!!! Thus fully decarburized surfaces are more susceptible than normal pearlitic/ferritic structured materials. However, I'm not suggesting that having harder HAZs represents any plausible defense against caustic cracking as residual stress is the big enemy to minimize and harder (than the basemetal) HAZs have been known to crack (driven by residual stress).

Bob.

 

[vesselguy]

bob330,
I understand from reading your previous thread why you would let the repair of a C.S. vessel weld in caustic service left un-PWHT. I said I understand why you did it, but I don't necessary agree with it. Today, we know about the mechanism and the effects of SCC with C.S plate products in sour and caustic environment. I am only suggesting the we engineering types should look at a piece of work as something that we will have to work next to for the next 10 years. If our decisions are sketchy, would we want to work next to that equipment? If not, then we know the answer to what we have to do. Equipment cost is tiny compare to replacing a segment of a plant and cost of lives.

By the way, if you do PWHT the nozzle, I would not recommend "bulls-eye" PWHT as this can make matter worst.

 

[robsalv]

On the topic of welding and temperbeading, would a higher interpass temperature produce a less stressed weld?

Could you get to a point with a preheat, high interpass temperature, half temperbead, postheat, slow cool down pad weld procedure where you could confidently rely on the repair to not be prone to CSCC??

 

[metengr]

robsalv;
You will obtain some benefit regarding residual stress reduction. The question is will it be good enough to drop you below the threshold of residual tensile stress to avoid SCC? In my professional opinion, no.

 

[stanweld]

bob330,
Since the vessel was never post weld heat treated and had been operating successfully for 40 years with no sign of stress corrosion cracking, it would appear that the caustic concentration and temperature range was/is below that which could cause CSCC. Provided that the service conditions don't change, your Method 1 repair technique was done with due dilegence.

robsalv,
While it is theoretically possible to adequately (not fully) reduce stress through controled preheat, interpass temperature and weld technique/heat input, it would have to be confirmed through test on representative welded specimens. Like weldengr, I don't think that it could be done or in any event that one could guarantee that all required parameters would be met during actual repair welding.

 

[bob330]

Thanks to all of you guys,

We are going to do our first internal inspection at a one year time frame instead of the origianlly planned 1.5 to 2 year plan. This will give us at least somewhat improved piece of mind.

Although we have not really observed caustic SCC in this vessel, the temperature and concentration data charts say that we should have. I was told when I started this job that this is a miracle vessel as it should have already failed. I sometimes wonder if it possibly was given a PWHT after construction 40 years ago and simply not recorded. Regardless, we are going to inspect our weld repaired area ahead of schedule. Hopefully, if its gonna crack, it won't be cracked bad enough to fail within a one year period.


bob