Shell distortion due to welding of lifting trunnion

[jtseng123]

Dear all,

Our inspector found out the inside of shell was bumped out along the shape of the lifting trunnion, that is obvious distortion due to welding. Please see the attachment.
Is there any analysis, criteria or measurement to accept the outcome ? If rejecting it, what will be the technical reason ? As of today, vendor is unwilling to fix it. I am afraid during lifting, it may crack or more deform due to residual stress plus lifting load. Or due to loss of strength by the distortion. That area able to take the design pressure is also in doubt. But so far we have no tool to reject it.

Please advise your opinion.

 

[chicopee]

Looking at the sketch, I am more incline to believe that the trunnion was impacted instead of the welding being the issue.

 

[TGS4]

1) What is the Code of Construction?
2) Is the design thickness governed by internal pressure or external pressure?
3) Is there an external pressure condition?
4) Based on the answers to 1) through 3) above, what is the permissible fabrication tolerance from the Code of Construction?
5) Are there Owner's Specifications (or other contractual documents) that have requirements that are more stringent that would result in a lower fabrication tolerance?
6) Exactly what are the distortions? Exact measurements/perturbations from a perfect straight edge (longitudinally) and from a perfect chord (circumferentially)? How do these compare to the fabrication tolerances from above?
7) Is the vessel to receive a PWHT?
8) What service is the vessel in and is tasty service susceptible to a degradation mechanism that depends on high stress (either applied or residual)?
9) Why did no-one in the design phase notice that there might be a risk of shell distortion due to excessive weld? Why was a pad not employed at this location?

I have more questions, but really, you and your team should be asking (and answering) these. You may need expert-level assistance on this. Even if the fabrication tolerances are being respected, a fitness-for-service exercise may be necessary, one that considers the exact (as measured by high-fidelity laser scanning) deformation.

 

[IFRs]

Yep - thin shell material + heavy weld = distortion. Not too surprised. Can it be load tested? How much weight is being lifted? What is the thickness and grade of material in the shell? What was the weld size and electrode used? What NDE was performed inside and outside?

 

[DriveMeNuts]

In the photo, is the outer distortion ring where the trunnion re-pad is welded to the shell?
How thin is the shell?

 

[MJCronin]

If this is indeed a thin-shelled vessel where welding of a stiff trunion COULD BE EXPECTED to cause weld distortion

Then why is there not a reinforcing pad to distribute the load and minimize the amount of weld distortion ?

Because of this thick/thin transition, I also suspect that there was no detailed evaluation of the stresses generated in the shell when the vessel is lifted.

I believe that this was a total "slapped together" design where the lifting trunion from another (thick walled) vessel was slapped on the thin-walled design

Is this true ?

MJCronin
Sr. Process Engineer

 

[jtseng123]

All,
I am providing more details as shown in the attachment. What is your recommendation of action ? Remove the entire region including re-pad and patch with a new insert plate with trunnion ?

 

[TGS4]

My recommended course of action is to start by answering a bunch of questions posted above.

 

[jtseng123]

TGS4,
This is just a simple 304 SS vessel with Div 1. No PWHT, no external pressure and nothing special. The thickness for all components are reasonable. Vendor's WPS/PQR were approved by our experienced welding engineer, but it does not mean anything because it is still human to operate, and depends on the skill, mood and whether he will follow the procedures strictly to do it right. We can't have 10 inspectors to watch 10 welders who welding a vessel at the same time. Fabrication mistake/errors always happen through my decades in this industry.

I am trying to gather some technical assessments to reject it and force vendor to re-fabricate. Otherwise, as one of our guys said, vendor can grind flush it and telling us it is fixed and then ship. Who knows what will happen during lifting and during operating. We have internal conclusion that vendor must be overheating it. But I am looking for guidelines if any that it shall be rejected due to such and such reasons, or a good guideline how vendor can fix it without gouging out the entire lifting trunnion and replacing with new. We also have a deadline to ship from Asia to US very soon.

 

[TGS4]

1) What is the Code of Construction?
2) Is the design thickness governed by internal pressure or external pressure?
3) Is there an external pressure condition?
4) Based on the answers to 1) through 3) above, what is the permissible fabrication tolerance from the Code of Construction?
5) Are there Owner's Specifications (or other contractual documents) that have requirements that are more stringent that would result in a lower fabrication tolerance?
6) Exactly what are the distortions? Exact measurements/perturbations from a perfect straight edge (longitudinally) and from a perfect chord (circumferentially)? How do these compare to the fabrication tolerances from above?
7) Is the vessel to receive a PWHT?
8) What service is the vessel in and is tasty service susceptible to a degradation mechanism that depends on high stress (either applied or residual)?
9) Why did no-one in the design phase notice that there might be a risk of shell distortion due to excessive weld? Why was a pad not employed at this location?

I think that you answered 1), sort-of 2), 3), and 7). I can't quite tell, but it appears that there was a repad, but the trunnion-to-pad weld seems to have been made to connect the trunnion to the shell. In this sort of configuration, I would have used a solid (not donut) repad, with the trunnion welded to the repad only. Then weld the repad to the shell. This smells of design issues, not just fabrication.

But, nevertheless, if the deformations are within fabrication tolerances, you don't really have a leg to stand on to reject the welding, do you?

 

[r6155]

Wrong contract.
Where is the inspection plan?

Regards
r6155

 

[DekDee]

Very poor design which appears to have been approved by you, the client so you are basically stuck with what you have got.
Why was the repad not a solid plate - as TGS4 has noted ?
If there was no "donut" on the pad there would not have been any distortion.

 

[jtseng123]

Welding lifting trunnion on top of re-pad is not a common design for many reasons: you can not take the combined shell and re-pad thickness for local stress analysis that ends up re-pad can be too thick, may exceed 1.5x shell thickness which is not preferred. Slippage occurs between re-pad and shell, re-pad may be warped during lifting, etc.
Trunnion on top of re-pad is mostly used for light load, can also be seen on pipe/platform support clips. In this case, vendor must check the stress on re-pad only first, making sure it is thick and wide enough, then check the shell stress at the edge of re-pad.

For medium to heavy load, commonly design is trunnion welded on thick insert plate, or trunnion welded through re-pad to shell to make it an integral part to resist lifting load.

As of today, vendor wants to perform FEA. We reject it because it has no meaning to do so without knowing if the mechanical strength has been changed by overheating and the shell protrusion, and how much residual stress is there. We are also afraid it may crack during hydrotest or during lifting.

 

[david339933]

Something isn't adding up here.....trunnions look like they are designed for 300 - 400 tons, yet vessel is 10.5ft ID x 18mm thick?

 

[MrPDes]

Often large trunnions are chosen to spread the load on thin shells rather than to lift big loads.

I've used the same size and design of trunnion on a 5000mm ID x 10mm carbon steel shell and there was no distortion. The welder must have really blasted it with allot of heat. My vessel was at full vacuum during transport and operation, so internal stiffening rings were needed.

I doubt the distortion is great enough to affect the primary stresses during the lift.

During operation, the area of the shell under the Trunnion will experience stresses of about half of the stresses in the shell away from the trunnion as the least stressed area on the entire pressure vessel is under that hulking big reinforced Trunnion, so stresses during operation are unlikely to be a concern.

If the vessel cycles from zero to MAWP, the stresses at the trunnion-shell junction will cycle from zero to approximately half of MAWP, so fatigue is unlikely to be an issue.

I often visit a plant with a vessel that has a big dent in the side of it because the vendors Engineering team didn't design the transport saddles correctly. 20 years later and that dent is still there.

It is disappointing when this sort of stuff up happens, however is this distortion which is out of site and where primary stresses are low really a big problem?

 

[r6155]

My proposal
1)100% Penetrant test, internal and external surface of shell in the area of lifting trunnions and welds in trunnions.
2)Hydrostatic pressure test with Acoustic Emission Test.
If 1) and 2) are OK this pressure vessel is OK.

Regards
r6155

 

[DekDee]

27 mm thick trunnion welded full penetration to 18 mm shell
20 mm thick repad then welded full penetration to trunnion/shell.
That is a lot of weld in one area - hardly surprising it has distorted.

Did the Detail A have a note for the trunnion to be welded first and then repad installed after or did they just weld both joints at the same time ?

 

[jtseng123]

r6155, all tastings are mandatory when it comes to lifting device. This is our requirements on every project. Our construction team (we are an EPC company) will not lift anything if it is not 100% MT/PT or even UT for welds and base metal. But can't wait until hyrotest to prove it is good. It is 50/50. No risk is allowed. If failure in hydrotest, tt can miss the boat and delay our construction.

DekDee, we will back check if vendor can reduce the thickness and increase trunnion diameter to cut down the heat input when re-fabricated, and request vendor to have QA/QC watch out how welder doing the weld. We already noticed the 20 mm fillet weld inside the trunnion is too big. 10 mm is enough.

 

[CuMo]

Is it not possible to replace just a section of the vessel where the trunnions are located?
Still better than replacing the whole vessel.

 

[r6155]

My second proposal
1) Cancel actual trunnions. Cut them close to shell as appropiate
2) design new trunnions and install them accordingly.

Regards
r6155