Pipe to Flange Design Question

[doezer]

Hi all.

i have very limited experiecne in weld design, and have been assigned a task to look at the design of a flange/pipe welded connection.
im reading all over that a full penetration butt weld is the best.
But why is that?

When I look at a socket weld, the weld seems to be so much more supported, structurally I mean.
You have essentially an inner sleeve located within an outer sleeve. So any structural loads occuring (bending moments etc) are not being taken directly by the wleded joint.
Whereas with a butt weld, its fully 'bendable' at the joint.

I understand you are creating a crack, but if you can verify that the crack is ok for the duty cycle involved and the fluids and pressure involved,
then surely you have a stronger joint structurally??

by the way, this pipe/flange joint will see very high operational loads, not just pressure load, also structural loads due to operation.
the pipe will vibrate and apply bending whilst the flange is secured, so the weld needs to withstand it etc.

hope this makes sense? as I say am new to welded joint design

thanks
D

 

[BigInch]

The full pen weld is the strongest. More stiffness than needed is not necessarily a good thing.

 

[doezer]

Secondly....

are there some online design guides available (youtube or otherwise) which can get a guy started on this...

i.e. that cover pipe weld joint design... i can find a whole lot on basic weld joints (plates, fillets etc),
but specific to pipes not so much
Tnx

 

[europipe]

You must not look from a structural view to pipelines, the load on a weld must be as small as you can get, it's only for connecting components. A socket weld has a gap where corrosion can occur, and socket welds are only for small sizes.

 

[doezer]

@ BigInch

Thanks a lot.

But in my case I have to allow for quite a large loading occuring which is not due to pressure, it will be in the form of a bending load on the pipe, tearing it away from the flange so to speak.
so wouldnt a supported 'sleeve' type weld be better here. intuitively this makes sense to me.

but i woudl agree that everyehre i look its saying use a butt weld.! But is this more for pipes with very predictable pressure loads only?

 

[doezer]

thanks europipe..

i have an additional constraint.
the material being welded is only good for welds to a depth of 3mm.
the pipe wall thickness is larger than that (~4 or 5mm).

so a full penetration butt weld is not a possibility here.
So this is why i am thinking a support sleeve located inside the pipe ID, with a very tight tolerance.

it will support teh weld and ensure only axial loads are applied to the weld.

OK, this all creates a definite crack in there, but its very quanitifable and can be analysed versus the duty and fluid compatibility etc.

would that make sense to you??

i cant see another way around that IMO..

d

 

[europipe]

Let a stress engineer look at the situation.

 

[LittleInch]

You appear to have a special situation as I'm struggling to think of a material which only allows welding to 3mmdepth

also a "support sleeve inside the pipe ID" is also quite odd.

Draw up what you're thinking of, post it, tell us what this material is and we might get somewhere, but at the moment all we're doing is blindly thrashing about in the dark and you're drip feeding important information.

your initial statement "So any structural loads occuring (bending moments etc) are not being taken directly by the wleded joint." Errr yes it is. You also get a high stress concentration at the weld and the tendency is for either the weld to crack or the pipe to buckle at the end of the socket weld.

Butt welding the pipes essentially means you have a homogenous pipe.

"by the way, this pipe/flange joint will see very high operational loads, not just pressure load, also structural loads due to operation.
the pipe will vibrate and apply bending whilst the flange is secured, so the weld needs to withstand it "

This is quite an unusual requirement requiring much more than weld design issues. The weld may need to be of high quality and all the issues you present would go into something like an Engineering Criticality Assessment (ECA) to determine acceptable weld defects.

Everything else is stress related.

Socket weld is only for small bore connections. Small bore connections are notorious for failing in service, especially if subject to high stress, vibration and static loads.

Be very careful.



Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[doezer]

Ok Thanks to LittleInch and all..

sorry. I do realise im drip feeding info here.

i will attempt one last post with some info, and if its still not enough, i will have to go away and figure it out some more before i annoy you all any more

The material is Al6061-T6. i have inherited the job from someone who left and so im looking through a data file that's pretty incomplete here..
application is aerospace - its a feasability study.
We have some data on file from an established weld supplier, and they have trialled some welds for us and established 3mm as a max depth for this material when welded (for a certain diameter)
Beyond that they get porosity to unacceptable levels for this application.

yes, i would agree with previous posters that the problem seems to be more a stress one than a piping one.

one solution is to use a slip on flange with a double fillet weld, one top and one bottom
so then there are no artificial cracks generated.
i understand the welds are weaker in terms of fatigue etc, but if the fit between the flange and pipe is very tight, then this should react loads, not just the weld.

d

 

[doezer]

sorry. one more point.

you say this is a very unusual requirement. the fact that we have high opertaional loads and high pressure loads.

would it not be quite normal to have a pipe/flange joint which is subjected to high internal pressures,
but also at the same time the entire joint is being subjected to further vibration loads etc.
in this case we have a propulsion engine, the pipe joins two separate components of the engine which vibrate and load is transmitted across through the joints,
but im thining this could apply for lots of other applications too, no?? (automoticve, aircraft, etc). ?

 

[BigInch]

Not completing a pipe wall weld to full penetration depth, is known as undercutting. It introduces stress concentrations which may result in thermal cracking even as the weld cools, if not at a later date due to thermal cycling (hot/cold environments that aerospace is famous for), is conducive to cracking, which is condusive to vibrational fatigue, which is conducive to corrosion... stay away from it. For us pipers in chemical plant service this is bad enough. I sure wouldn't want it on an aerospace app.

You may need to do a welded joint on what remains after making a machined internal bevel. Spend some time on it. Keep it safe.

 

[Compositepro]

Welding Al 6060-T6 is a completely different story than welding steel pipe. The weld will completely remove the heat treatment and will be considerably weaker than the parent material. Are you planning to re-heat treat the whole assembly after welding?
I can't recall ever seeing flanged pipes used in aerospace applications. It is a relatively heavy type of connection.

 

[LittleInch]

The joining of a pipe - relatively thin wall, quite flexible to a flange - thick wall, rigid, induces considerable stress impacts, especially at the joint / weld and in the local pipe region next to the flange.

without doing some fancy FEA analysis - this might be required in your case - in general flanges and pipes just sit there under internal pressure.

Additional loads and especially vibration causes many problems and much angst with stress engineers

Anything you can do to avoid stress on the welded joint is better, hence weld end flanges.

All others have large potential for weld cracking and high stress concentrations.

Lots of locations like this will use flexible pipe instead of rigid pipe.

According to mr google Al6061-t6 has good weldability...

for me choose a different material.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[SwinnyGG]

If your welding contractor is telling you they can only apply a 3mm deep weld on 6061-T6, you need to find a new welding contractor.

 

[doezer]

Thanks to all.

im new to this forum and i have to say its really great to connect with other engineers with different beckgrounds like this...really helps so much
I hope to be able to contribute and not just learn on at least some of the forums at some stage down the line..

so the overall perception is that this is a fairly demanding application, not the norm.
pipe/flange joints are usually used in situations where they just sit there and do their job (presure containment)
I may be best re-posting this one over on stress forum?? i.e. ways to protect a pipe/flange weld from high loads...

i also have the added problem that the solution will have to be acceptable at ESA (space agency) levels..
a question though! would it be safe to say that socket weld arrangements, or slip-on flange weld arrangements have higher structural load resistance?
although the weld is working in shear there isnt it?

@ jgKRI
theres a nice curveball... if the depth of 3mm deep is not a limitation that changes things for sure..
one thing they are saying to is us yes you can weld to deeper than 3mm or so, but you will get porosity levels that are out of spec foryour application...
could i PM you about that please??

thanks
D

 

[BigInch]

The degree to which socket welds and SOs are limited to small sizes or outright prohibited in the petrochemical industry would suggest they have little advantage, if any.

 

[LittleInch]

"would it be safe to say ..." Not in my opinion, no

one thing that is good here is the search facility at the top of th epage.

learn how to use it well and you find things like this

http://www.eng-tips.com/viewthread.cfm?qid=185899

For simple, low stress, low bending small bore connections, socket weld is acceptable.

As soon as you put vibration or high stress other than hoop stress then they have a history of failure

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[doezer]

ok folks. some developments.

If i compare the hoop stress (PD/2t) in the butt weld for a pressure of 200bar, OD 69mm, thickness 3.3mm
I get a stress in the weld of 134MPa (= 19ksi)

If I change that to an end weld, so just a face weld, welding the OD of the pipe to the flange face,
I get a shear stress in that end weld of just 13.2MPa (= 1.8ksi). So its 10 times less.

So the face weld on the end weld is a lot more desirable there. OK the shear allowable is alot less but even allowing for that the
stress in the weld is way way lower than in the butt weld.
So why is the butt weld commonly considered the much better option by everyone?.

(these are assuming a full pen butt weld of wall thickness 3.3mm vs a end face weld of depth 3mm)

 

[BigInch]

Can you post a diagram of this and your calculations if possible?

 

[doezer]

ok here goes. i feel like im submitting an exam paper, go easy on me

d