Shear Lugs (Axial Stoppers) on Pipelines 1

[TomoB]

Hello,

I need to check the shear lugs (axial stoppers) on pipelines with respect to the loads that were calculated in the stress analysis. I came across a project where Finite Element Analysis was used for these components. But is this really necessary?

What is your approach? Do you rely on hand calculations, and if so, which methods or standards do you use? Are there specific procedures or rules of thumb that you follow for designing these components?

Thank you for your time!

 

[LittleInch]

I've never seen anything like this for no doubt very good reasons.

These are just set up to rip a hole in your pipe.

If you want to line stop a pipe, your either clamp it very tight or weld it circumferentially or a long axial weld (not preferred). Those little tabs are the worst of everything.

you sometimes see trunnions or similar, but not those little tabs, sorry.

 

[Snickster]

A lug such as that is commonly used for limit stops welded to pipe bearing onto structural support to stop axial movement of the pipe. It is similar to what you are doing instead bearing against clamps they bear against the structural steel which stops the line from moving.

I would be concerned about using clamps like that since you are not sure if the axial load of the pipe trying to move will be evenly distributed among all lugs that bear on the clamp causing higher local stress in any given lug. Also the clamp is not specifically designed for point load application.

That being said prior to computers the manual method used was WRC 107 which is very tedious but gets you results you need. There are other quicker methods to estimate. I will check my notes for examples and post later.

 

[Snickster]

One of the easiest methods for calculation of local stress due to attachments on piping is the Kellogg method. This is what really old timers used to use prior to computers. It is proven to work though although may be a little on the conservative side. In fact companies I worked with recently still use it as a first check. If the stress passes base on Kellogg method no further analysis is done, if not then use WRC 107 or FEA.

Local_Kellogg | calcstress

www.calcstress.com

 

[LittleInch]

Lugs tend to have reinforcement pads.

That clamp could also move suddenly leading to high shock loading.

 

[Snickster]

The kellogg method has reinforcing pad option. Personally I would not use that configuration for an anchor with very high loading. If that is a sliding support that it is attached to maybe, but probably not as you don't know how the clamp will react to that type of loading, but I would at least design every lug to take the entire load assuming that any lug will be resisting the entire axial force.

If I were trying to stop axial movement like that I would use W shape beams installed in vertical on each side and horizontal at top and bottom forming a box shape where the lugs would bear against.

 

[Snickster]

Here is some details of lugs attached to pipes to form stops.

 

[TomoB]

@LittleInch I did not invent this, this is a concept that was used on the projects before for HP Steam systems under PED III category. So this was also approved by the Notified Body in the past. Other than if you take a look at many spring manufacturers the vertical spring hangers can be attached by either shear lugs or by trunnions. So I would say this is nothing new.

@Snickster thank you for the insight and the documents, I will check them.
For sure the idea was not to count all 4 shear lugs as a bearing, but to account for distribution and assume that maximum 2 lugs will carry the load. I have seen before some guidelines (but cannot find the source now) for this design - if 2 lug configuration is used, 1 takes full load, if 4 lug configuration is used, load is distributed to 2 lugs.

 

[LittleInch]

It might be nothing new, but still looks like a poor choice to me when other, much more suitable connections are available.

A lot does depend on the force you're expecting to see, but I still wouldn't do it. Too much risk / uncertainty for very little benefit. IMHO.

 

[TomoB]

Thanks, I will probably still have to go with these ones as the support concept is already made this way and the project is in the late phase.
The other limitation will be if the Pipe shoe supplier will confirm this loads.
One info - LISEGA provides pipe shoes with trunnion insert for axial slipping prevention, and this would be the best solution here. But have to go with the lugs for now.

 

[LittleInch]

Can't see why you couldn't just convert this to a circumferential strip of metal the thickness of the clamp and then do a full circumferential weld on at least one face.

Just becasue it's late in the day doesn't mean you need to continue with it.

And from your OP, you still need to sign these off or "certify" them in some way?

If it's thick enough pipe (> 12.7mm), then you're prob ok, but anything thinner and it looks higher risk to me.

 

[XL83NL]

Not that it helps much (for now), but ASME is working on B31W which will cover the design criteria for sizing such lugs/attachment. I think it's reaching a final stage now, so hopefully it's released soon.

The LISEGA option is interesting, but I'm not sure what assumptions they've made in designing these. The load bearing/resisting capacity of these items may depend on how much the pipe is already stressed due to sustained and/or thermal displacement stresses.

 

[DavidCR]

Lugs or clips can be calculated with formulas by Mohs or Kellog, and check stress on pipe with WRC 107.

I´d use PVELite software lifting lugs analysis that include all that ( impact and ocassional loads factors can be considered). Ideally an analysis with NozzlePro software would be perfect for the issue.

If you use pads (as big and thick as reasonable) maybe stress pipe would not be neccesary depending on various variables (loads, pressure, thicknes, diameter).

I agree is not the best option but not a terrible practice.

Big pads or a complete welded ring band would work similar and stress analysys would not be advisable if it is thicker than the clamp.