Complete penetration welds

[MotorCity]

I usually don't give much thought to complete penetration welds since they develop the full strength of the base material. However, I am wondering if there are reasons to specify a particular type of CJP weld....double bevel, double V, single V, square end, etc. We usually leave it up to the fabricator. For reasons unknown to me, I understand that prequalified welds may require a particular root shape and gap. I guess my question is, if they all develop the full strength of the base material, why not just use the weld type that involves the least amount of weld material all the time (whichever one that is)?

 

[SJBombero]

The type of CJP is determined by considerations of economy, access and edge geometry. For some connections, any type will work but some would obviously be more economical than others. Some fabricators have processes setup that would make them prefer one type over another. I have read structural engineering journals claiming that specifying a blank weld symbol with just a CJP tail is lazy and bad practice. Unless there is some sort of consideration such as lamellar tearing that needs to be considered, I think that that practice could be justified at least some of the time.

 

[phamENG]

MotorCity - my training and thoughts match yours pretty closely.

To the tail with 'CJP' on it - I agree that it's justified sometimes. If you are doing the connection design, then it is not justified - you should select the appropriate weld and specify it. If you want to put a line in your notes/specs saying that the fabricator can propose alternate welding procedures, then you can...but at least there a baseline for everyone to bid to. If you're delegating your connections, then I think the CJP on the tail is a good way of communicating to the SSE that you want/expect a CJP in that location; the design of which is up to him/her.

 

[KootK]

Like SJB, I suspect that the various options are indeed mostly about shop efficiency and sconstructability. Available processes, different access conditions etc if field welding.

The only engineering consequence that I could think of would be related to dynamic & fatigue loads potentially. While full strength is developed across the depth of the weld, I would not expect the welds to be uniformly identical from a ductility perspective across the depth of the weld. That said, off of the top of my head, I don't recall any specific requirement in this regard.

 

[Flotsam7018]

I think leaving it up to the fabricator is typically the best option actually, or at least consulting the fabricator. Once the decision has been made that a CJP is required the type of joint won't typically alter the design approach, but the preparation to make that joint a reality can be quite different.

 

[dauwerda]

The only engineering consequence that I could think of would be related to dynamic & fatigue loads potentially. While full strength is developed across the depth of the weld, I would not expect the welds to be uniformly identical from a ductility perspective across the depth of the weld. That said, off of the top of my head, I don't recall any specific requirement in this regard.

The only thing I can think of that gets close to this (at least from a code perspective) is if it is a CJP weld that is made from one side and utilizes backing. The backing (and its attachment to the structure) can cause fatigue/ductility issues, which is why it is required to be removed in some seismic design applications.

To the tail with 'CJP' on it - I agree that it's justified sometimes. If you are doing the connection design, then it is not justified - you should select the appropriate weld and specify it.

I disagree. If a CJP weld is required and that is the only requirement you, as an engineer have, there is no reason to be dictating exactly how that CJP weld should be accomplished. Let the fabrication shops select their preference. If you are working directly with a specific shop and you know what their preference is, sure go ahead and put it on their, but if not, I don't see how it is helpful to be adding more restrictions than necessary.

I guess my question is, if they all develop the full strength of the base material, why not just use the weld type that involves the least amount of weld material all the time (whichever one that is)?

Weld economy is based on more than just the amount of weld metal deposited. Other major expenses that come into play:
- backgouging, which is required when performing a two sided CJP weld.
- edge preparation, a 2 sided weld requires more edge prep
- material handling, a 2 sided weld requires parts to be flipped over and handled more
- backing, a 1 sided weld will typically require the addition of backing

AISC's Design Guide 21 has a fantastic discussion on weld economy in chapter 17 (well the entire chapter is dedicated to just that)
It summarizes that, "single-sided welds are usually easier to prepare and make, making them preferable to double-sided welds when distortion control is not a concern"

That sentence also hits on one important topic - distortion control. A single sided weld will have a lot more material deposited on one side, which can/will cause excess distortion unless it is restrained or prevented.

 

[noratetoobig]

A few points to consider.

Although most engineers leave it to the fabricator who gen don't have engineering knowledge about a project (if at all). One should pay attention to full penetration welds where fatigue is concerned since welds shrink creating internal stresses. Generally for static design with relatively thin plates (gen up to 50mm) these can gen be ignored due to local yielding within a welded joint when loaded.

Gen fabricators will use single sided pen welds for thicknesses less than 20mm (some have a limit of 25mm). Also one has to consider how accurate (tolerances) the local structure needs to be once welded since plates deform from the welding process which can lead to a substandard product which then may need to be machined flat (as in a support) or a combination of machining and heat treatment to try and bring the local structure back to shape.

For large penetration welds the fabricator would often use double sided with a 60/40 or 70/30 split on weld. and remember any plate with a full pen weld uses twice the volume of weld as for a double sided pen weld with the associated shrinkage. And for really thick plates 50mm+ supporting critical loads, post heat treatment may be needed to reduce the inbuilt shrinkage stresses.

This is also linked to the materials being used such at a cruciform joint with incoming plates welded with full pen welds and which have tensile loads applied, Z quality plate with improved through thickness properties should be used. (or the material assessed for low sulfur and good ductility).

That's my two pennyworth. However, each case may need to be assessed depending on type of structure and loads applied.