Pipe to plate weld selection

[sekwahrovert]

I frequently design monopole-type structures with pipe-to-pipe splices subject to large reversible shear and moment loads from wind. I've attached a sketch of a typical connection with 2 options for the weld of the plate to the pipe. I'm looking for opinions: which is the better weld for this connection?

 

[KootK]

I would expect superior performance from option one. I suppose the real question is when is option two good enough as it looks more economical.

For me, it would come down to my expectations of joint behaviour. If the flange plates were flexible and resulted in significant bending moments being transferred across the flange plate to pipe joint, I would go with option one as the lever arm between the two weld is very small in option two. Alternately, if the flange plates were very stiff and the welds would only see direct shear stress, I would go with option two as a result of economic considerations.

Every time that I go snowboarding, I wonder how these connections are designed, especially when stiffeners are involved. Is there not a design guide for this out there some place? I've got a book on tubular steel structure by Troitsky that might have something. I'll check it out when I get home tonight.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[dhengr]

Sekwahrovert:
This is a problem that would probably be interesting to study a bit with FEA. I would like to study how the loads and stresses are distributed and flow btwn. the bolts, plates and pipes, with the slightly different arrangements. Understanding that welds and sharp changes in direction always show hot spots, but it’s the general stress flow and stress level that I’d like to study. Otherwise, some food for thought off the top of my head....
1. I would want to know more about your fabricating equipment and methods because they will play into the difficulties of each detail. The right beveling and weld prep. equipment makes detail #1 pretty clean and easy. A roller system is a must for pipe girth welds and automation certainly improves productivity and quality. I don’t much like backer bars, they are a necessary evil in many instances. On big enough pipe, I’d sooner back gouge and weld from both sides, with a reinforcing fillets.
2. Detail #1 is easy to fit-up becuase the i.d’s. of the pls. is at least several inches smaller than their mating pipe, maybe they are solid pls., not annular. They will tend to distribute stresses around the pipe at any concentrations, because of their larger size, (width, o.d. minus i.d.)/2.
3. Detail #2 may be a bit of a fit-up problem as a function o.d. and round vs. oval tolerances of your pipe supply. I’d want to study that a bit, and I’d like nice snug fit (no big gaps) at the fillet roots. You also have two different welding set-ups for the two fillet welds. And, I’m pretty sure more filler metal and weld time for the same weld cap’y. This detail takes end pl. loads and stresses directly into the pipe, and has some (causes some) tensile stresses perpendicular to and across the root of the outside fillet, not a good condition. You will most certainly want a larger fillet for the outside weld. The inside fillet is fairly much less loaded.
4. Larger differences in pipe o.d. will complicate the analysis and should be studied with the FEA work. Do the same study, varying only that o.d. difference, to see what that does.
5. Stiffeners are a tried and proven reinforcement for some of these joints, but they do take the loads from the end pls. directly up into the pipe, and that’s good. But, they also have their own complexities and stress concentrations. A thicker end pl. is certainly cheaper than a bunch of stiffeners and their application.
6. This is probably one of those details that we Structural Engineers like to beat to death, without ever coming to a certain conclusion. Keep it clean, of high quality workmanship, and then whichever works best for your shop is probably pretty good.

 

[hokie66]

Option 1. I don't know why...just the way I have done it, and seen it done many times.

 

[TLHS]

Detail 1 with a solid plate is obviously the best option. Your flange plate is likely significantly stiffer than your main structure wall. Instead of the load path involving a reasonable amount of bending into the main member wall you can start assuming that the flange plate is basically simply supported by the both the near and far member walls.

You can sort of do the same thing with a stiff annular plate as well (relating to the size of hole and thickness of plate), but the analysis and load path are a bit more complicated. You also start seeing a lot more deflection.

I actually did some fea investigation of this a few years ago when I was trying to get creative on some pile caps plates, but can't recall enough to give you more than the basic impressions above.

 

[sekwahrovert]

Thanks for all of the input. Kootk, I understand what you mean about bending moments in the plate being transferred through the welds. However, I'm not sure how the option 1 is more effective at transferring the bending stress. As you say, the lever arm is very small in option 2, but it is even smaller in option 1 if you consider the distance from the backer plate to the reinforcing fillet on the CJP. If the plate is not rigid, how do you picture the bending moment being transferred through the CJP?

We have used option 2 for quite some time with good results. The weld is being questioned by a third party, which is why I'm looking for input. A couple of resources that are helpful in designing the plates and bolts but not the welds:

Design of Monopole Bases by Daniel Horn
Design Guidelines for Annular Base Plates by Ronald A. Cook for Florida DOT

 

[SAIL3]

ASME code has some guidelines for fla and bolt design....

 

[carnahanad]

Sekwahrovert,

Is there really a bending force in Option 1? The backer bar is there to only get the full pen weld started between the two base metals. If fabricated clean enough and shop welded, a backer bar may not even be required. I don't believe you can count on the backer bar creating a moment unless it has been designed to transfer the load through itself and through a separate weld to the base plate. However in option 2, in order to get both welds to see stress, there has to be a moment in the base of the pipe which could cause problems. Also, I'm not sure for this type of construction, but backer bars must be removed from seismic connections (for building design) to avoid force transfer through the backer bar that isn't expected.

In option 1, I think you could make the assumption that the whole Pipe is resisting the main "M" moment so the welded connection should only see the couple force T/C reaction creating vertical shear in the weld as well as the horizontal shear from "V". It seems from your sketch that Option 1 would take care of downward P from bearing alone.

In option 2, both the V and M get transferred to horizontal shear and the P value is vertical shear in the welds.

 

[KootK]

@OP: I guess I was assuming that option one would always be a continuous circular plate. If both options are annular then I agree, the weld bending issues are similar.

I've done some utility pole designs that were based on very rigid municipal standards. Always your option two.

And thanks for the FDOT info. I was previously unaware of that. For what it's worth, I consulted my Troitsky book and found no silver bullet there. Hopefully the ASME docs yield more fruit.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[JeanLucPicard]

Just yesterday, I've designed a similar connection, with option 1. I never seen (than again I don't have much experience) option 2. After examining op2, I would rather go with opt1. Because, of the weld work that is doubled in the opt2. Skilled welding hand is very very expensive...

Live long and prosper!

 

[SAIL3]

At first glance the welds should only experience the loads delivered by the pipe, however if the the fla is flexible and one experiences dishing of the fla, then , secondary stresses from this deflection could be significant as it works it's way back into the pipe especially if the pipe has relatively thick walls...again the ASME code has developed guidelines for fla design , if this is a concern...

 

[KootK]

I'd like to check out the ASME stuff Sail. Could you point me in the right direction? Is there just one ASME code? I don't play in that sandbox.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[SAIL3]

Koot, if you just want to get an overall familiarity with the ASME code, I would sugest a few books that may be helpful since the code itself is large and takes alot of time and effort to become familiar with.
I am not an expert on high pressure vessel design, but use portions of the code in my srtuctural design.

Pressure Vessel Design Manual by Dennis R Moss
a similar and very helpful book by Bendar which I do not have and may be out of print , but an excellent book.....

Coming back to the OP's problem...I guess ,with me ,it comes down to a few basic questions...
if I can reasonably assume that the fla is pinned @ the pipe, then option 2 would be fine with me.
if, however, the local stiffeness of the pipe wall(wall thickeness) is becoming significant, then I would be less comfortable with option 2. Ofcourse, the relative stiffeness of the fla can mitigate these concerns.
Using ASME guidlines for fla design by hand can be a PITA .

 

[SAIL3]

correction...that should be Pressure Vessel Design Handbook by Henry H Bednar...

 

[KootK]

Thanks Sail. I'm presently only interested only in the ASME provisions that address this situation in particular. Can you be any more specific?

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.