Groove weld gap for re-pad

[jtseng123]

All,
As attached sketch, to consider combined thickness of (shell + pad) to check the local stress at the trunnion junction, what will be the groove weld gap as shown related to shell, pad or trunnion thickness?

 

[SnTMan]

jtseng123, I'd think an 1/8" would be pretty typical. Not sure it matters much if properly welded out...

Regards,

Mike

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[TGS4]

Does your calculation assume that the trunnion is fully attached to the shell?

 

[weldtek]

We typically weld the trunnion to the shell with whatever weld size is required. Normally it's full pen thru the trunnion wall and a fillet. When that's complete drop the repad over the trunnion and weld that joint. The repad opening needs to be sized to accommodate the previously applied fillet.

 

[jtseng123]

The issue is the use of combined thickness n WRC107 local stress analysis. If the trunnion has a huge pull out force, it may be very easy to separate the pad from shell if only 1 /8" groove weld for pad to shell, and I saw people using combined thickness in the local stress analysis, that does not sound right.

I also see other company's standard asking the groove gap to be equal to pad thickness, that makes sense such that pad is definitely integral with shell to share the load and the combined thickness can be used in the analysis. What is your experience ?

 

[Duwe6]

". . .weld the trunnion to the shell with whatever weld size is required. Normally it's full pen thru the trunnion wall and a fillet."-

Exactly - Full-Penn the trunnion, then lower the repad and weld it into the OD trunnion weld. Look at the repad detail dwgs for API 650 nozzle repads.

 

[jtseng123]

Duwe6, this is not a nozzle weld. There will be cases that trunnion will not be full penet to shell. It can be either a pipe support bracket that come with re-pad instead of trunnion with re-pad, and full penet is not necessary.

 

[Duwe6]

Have you ever watched a vessel being handled? The probability of the entire weight of that vessel ending up being supported by one trunnion is not trivial. It is a very real possibility. Assuming a good fitter/welder, the extra time/cost of full-penn [vs, whst you detailed] is trivial. Poor policy to skimp on highly critical items.

 

[dhengr]

Jtseng123:
The idea of a repad is to distribute the concentrated loading/stresses to a larger area of the shell. So, why would you work so hard to concentrate the stresses at the root of the trunnion, with a kinda trashy, unsymmetrical weld and repad detail? The proportions of your sketch seem to indicate excessive welding, which can be almost as bad as too little weld or an unsymmetrical detail. The weld btwn. the shell and trunnion does not have to be a full pen. groove weld if the loads and stresses do not req’r. it. But, it should be well detailed, well done, and defect free, becuase of the high loads it may see, the out-of-plane loading it may see, and the potential of impact loading during lifting. I would weld the trunnion groove to the shell without any appreciable reinforcement fillets, as mentioned above. Then slip the slotted repad over the trunnion, with a partial pen. groove bevel on the slot which would distribute 50-60-70% of the trunnion design load, how conservative do you need to be? The intent being to distribute some %age of the load further out into the shell. Then fill the groove btwn. the trunnion and the replate, and add a reinforcing fillet on the top of that, and a fillet all around btwn. the repad and the tank shell. Do a good job of welding the root pass of this groove weld. Round the corners of the repad so those fillets can more likely be clean continuos welds. The most critical weld location is at the corners of the plate edges on the trunnion, where welding around those four corners tends to cause a stress raiser, an undercut, right at the corner and the toe of the fillet. The easiest fix for this is to grind a slight radius of the trunnion plate edge/corners where the fillet wraps around the corner, this tends to prevent the weld puddle from sloughing down at the sharp edge/corner.

 

[jtseng123]

We have done fillet welds as shown on my original sketch if not thousands but hundreds of times for pipe support brackets，platform support brackets and other support brackets, etc., the same as any other engineering company is doing. Many times re-pads are required to reduce the stress at the junction, but is lack of a guideline how the weld shall be designed effectively and economically without using full penetration. The key will be the weld gap size to transfer the loads. Without going to FEA for hundreds of clips, I am seeking for people who have experience in doing similar design and advise a formula or engineering judgment in related to shell, re-pad or lug thickness for the weld gap width. Additional sketch attached.