Base Elbow - Pad Reinforcement vs. Heavier Wall

[StressGuy]

In piping systems it is often the case that an elbow near grade in the vertical orientation is supported directly from the ground by a base support/dummy leg welded directly to the elbow. A control valve station is a typical example. The attachment is usually another piece of pipe, but for the masochists out there, can also be a structural shape.

Many times, the local stresses at the attatchment are such that reinforcement of the elbow is required. This can be done via adding an external pad or by increasing the wall thickness of the fitting.

I understand that when a pad is called for, it is usually cut out of another elbow, which essentially means you have to pay for two fittings.

It would seem that, for commonly available wall thicknesses, that calling for a heavier elbow would be the more economical solution.

The main disadvantage to an increased elbow wall thickness is that it is typically a variation from the pipe spec, so has to be specially noted on the piping drawing and, once it's in place, it would be difficult to check during a field walkdown to make sure the heavier elbow is in place. It would also make it more difficult for a future stress engineer to analyze a change to the system, such as a new tie in, as, in my experience, documents such as fabrication ISO's are typcially lost by the owner after the job is complete and the engineer has to rely on the spec to "rebuild" a new model. Not knowing about the heavier wall would lead to a lot of head scratching by a future engineer who is trying to figure out why the piping system hasn't collapsed.

Anyone have any thought's or guidelines on this one that they find useful?

 

[11echo]

In my 26 yr.s as a piping designer, I have NEVER seen a reinf. pad on a base Ell supt! I've seen them on branch connections, but that's it! I would think that because a dummy leg of the supt is normally the same wall thickness as the pipe, and the length is minimal compared to the piping, stress at this point would NOT be an issue. Also when you joint to different schedules together, you back taper the smaller I.D. component to match the larger, semi labor intensive, not the best situation... But I've seen it done once(to meet project schedule). My 2 cents!
...Mark

 

[Tarkjell]

I have seen trunnions welded to elbows on offshore platforms and as solutions in company pipe support standards. My opinion is that these designs lack good engineering judgement when used extensively. However, perhaps a pad reinforcement might be the only solution, but should then be the last choice.

Consider the following before adding a reinforcement pad:
- Shorten the trunnion length to reduce bending moment.
- Use trunnion size equal to elbow size.
- Use low friction sliding pad to reduce friction forces.
- Increase the trunnion wall thickness.
- Forget the elbow trunnion; consider a different support arrangement.

Regards
Tarkjell
- Check that you are designing for coincident loads, not maximum component loads from different load cases.
- Consider increasing elbow wall thickness (as you have described already)

 

[Tarkjell]

Sorry, only elbow trunnions together with reinforcement pads are not good design. Elbow trunnions as such are fine!

 

[StressGuy]

Thanks for all the responses. I do think that a heavier wall elbow is the better way to go. Pretty much anytime you have a control valve station, you are going to have these types of supports on either side. When one of the dummy legs is anchored to the ground, I find that local stresses at the attachment are almost always an issue.

And, all you have to do is walk through any refinery along the Houston ship channell and I guarantee that you'll be able to find pad reinforced dummy legs on elbows. They are actually not all that uncommon.

 

[MJCronin]

Stressguy,

There was a paper written about 15 years ago by engineers at Stone & Webster on the proper fabrication and stress analysis of elbow-trunions..... as you described. Two of the authors were a Mr Hankinson and a Mr. Budlong

With regard to modelling these components into a flexibility analysis (e.g. CAESAR-II)the people at COADE offer a standard method in thier examples manual. The elbow has been (obviously)stiffened and is no longer as flexible as an elbow without the trunion.

You have touched on an area where different engineering firms have taken different approaches. Some ban these types of supports, some require reinforcement,.... others simply "leave the issue to the support designers" ( the project MBA solution)

I would not use a thicker elbow, but would add/or realign supports so that the load taken by an un-reinforced elbow was "modest" ( say 300x pipe diameter)

My thgoughts only...

MJC

 

[11echo]

OK guys lets talk about this. A dummy leg supt. is normally around 75% the dia. of the pipe you are supporting, the wall thk.s are also the same, and you use a combination weld (bevel weld & fillet - where the weldment is equal to or greater that the pipe wall) to attach the dummy leg. And by my books the MAX length on your dummy leg is 36". In normal application there is NO need for a reinf. pad.
IF you were to use one, would you not start to get into overlapping heat effected zones of the attachment welds? ...are you going to post heat treat this piping system? ...(do you see dollar signs here?*G*)
A base ell supt. is normal used at piping manifolds, inlet/outlets to equip., & piping runs that are low but not positioned on normal pipe supt.s. I would suggest to you that if your stress programs or caluations indicate the requirement of a reinf. pad on a base ell supt. you have problems else where in your piping configeration and need to address them with OTHER means. My 2 cents. ...Mark

 

[11echo]

P.S. ...talking with some stress engr's, about this type of connection & failures, the wall of the pipe does NOT fail, but the weld is the normal culprit. So going to a higher schedule pipe wall would not be a fix. ...AGAIN, my 2 cents. Good Luck! ...Mark

 

[Tarkjell]

Mark

It would be interesting to know whether or not full penetration welds, as indicated in figs. 328.5.4A, B, C in ASME B31.3, were used for the failing trunnions your stress engineers are talking about.

Tarkjell