AISC_PJP Welds in Tension (?)

[Wanna_be_SE]

Recently I was doing some weld calc's using Spec section J, and discovered the Table J2.5 lists "tension" as a possibility for PJP weld joints. This made me scratch my head, bc it doesn't list tension as a check for CJP welds (bc CJP welds are governed by the base material) and that made sense but when I saw tension as a possibility for PJP welds, I don't understand why.

For most (if not all I think) weld calculations AISC basically always take a failure plane at the throat of weld (most conservative). Table J2.5 lists tension criteria for fillet welds as can be neglected, and I assume for the reason that weld rupture (tension failure) will never occur bc welds are typically (maybe always) stronger than base material, and the "path of least resistance" will always occur at the throat of weld - which is a shear failure, not tension. Therefore whether longitudinal or transverse loading on weld, taking a failure plane at the throat (AISC) will always give you a shear failure, so I beg the question why do PJP welds shown in Table J2.5 have tension listed? If anything, I would expect the CJP welds (butt welds specifically say in a plate) to have tension listed, but then again bc weld metal is typically always stronger than the base material, the base material would fail before the weld would (in a butt joint in a plate).

Someone mentioned that it could be based on the type of loading (perhaps bending, as opposed to shear or axial), but even in that scenario the weld failure plane that AISC tells you to take is always (to my knowledge) at the throat of weld, and again that's a shear failure. If you consider say an I-beam girder welded to an I-beam column, and lets say the connection has stiffeners and there's moment developed. The top flange would put that weld in tension under bending, however the failure that occurs IAW AISC is at the throat (shear failure), so I can not think of a situation that would warrant checking the tensile strength/stress in any weld. In this example I described, most likely it would be a fillet weld so Table J2.5 says to neglect tension, but lets assume (for the sake of argument) that the top flange weld is a corner weld (square groove with reinforcing fillet). That may classify as a PJP and therefore maybe you could argue check tension in weld....but that would go against the failure plane at throat, unless you said that failure plane at the throat was a rupture, and not a shear failure, but I've never seen that in any weld calcs I've done and with that being said you'd still have a shear failure at toe of weld - between weld and base material, so that would warrant almost like a block shear situation but in a weld (which just sounds crazy thinking about it). Maybe through some testing they've discovered there is a possibility and that's why it's listed, but kind of seems contradicting to some extent given the assumed weld area AISC always tells you to take.

Has anyone else ever wondered this, or am I the only one that honed in on the tension for PJP welds?

 

[Wanna_be_SE]

OK, so I correct myself...I just realized the whole "block shear idea" wouldn't happen (I think) ever bc once the weld fails at throat that connection locally is gone - so no other failure plane (I blame lack of sleep!).

But in any case, at my previous employer we had SHEAR ultimate strengths for weld filler metal, based on type of filler metal with certain base materials being used. And it was for the reasons above that failure in a weld will only always ever be a shear failure (at least that's to or was my understanding), and so again the possibility of checking tensile stress/strength in a weld seems unnecessary, in any situation (except for CJP welds for example in a butt weld in plate - but again that's controlled by base material).

I did just think of a situation, based off another forum I just read, and that would be a PJP intermittent weld for butt weld in plate....but intermittent welding for butt welds is apparently prohibited by AWS (and that's another thing I've never seen and sounds kind of crazy anyways).

So i'm still left with the question - why does AISC list tension for PJP welds?

- Dan

 

[ProgrammingPE]

Below is my understanding of the load types that are listed in Table J2.5:

The table suggest that fillet welds never have a tension or compression load normal to the weld axis (unlike for CJP and PJP welds). As shown in the picture above, when a tension load is transferred through the fillet weld, they are considering this to be a case of shear (even though there is shear stress at the face of the top plate and tensile stress at the face of the bottom plate).

Tension that is parallel to the weld axis can be ignored for CJP, PJP, and fillet welds. This occurs when you are lapping two plates in tension as shown above. The stress in the weld would be the tension divided by the area of the two plates including the weld, but Table J2.5 says the weld capacity is not decreased due to this tensile stress, so you don't need to check for this.



Structural Engineering Software:

http://www.structuralcentral.com

Structural Engineering Videos:

http://www.youtube.com/structuralcentral

 

[Wanna_be_SE]

ProgrammingPE (Structural),

"A picture is worth a thousand words" - thanks that clears things up a bit for sure.

With your sketch of Tension normal to weld axis, I can now understand what they may be talking about...however, I think the failure plane would still be at throat of that bevel filled with weld. AISC lists allowable stress for weld (with Tension normal to weld axis) as 60% of Fexx, therefore they are still considering a shear failure for that type of loading. The only "shear failure" I see in that scenario is at the throat of bevel as described. The interesting thing is for this specific scenario the strength reduction factor (LRFD) is 0.8 instead of 0.75, so AISC is allowing you to go higher - which still doesn't make sense bc it's a SHEAR failure, not tension. if I was doing this calc, I would use 0.75 and take shear plane at throat of bevel (as I would for pretty much all weld calcs).

With respect to the CONNECTION, I'd have to disagree with your sketch for the Fillet tension parallel, bc yes the stress is P/A but for sizing fillet weld I would still take a failure plane at throat and do my weld calc based on that (each fillet resisting P/2).

 

[ProgrammingPE]

however, I think the failure plane would still be at throat of that bevel filled with weld.

The failure plane for the PJP weld is at the face of the bottom plate as shown in the image below, which comes from AISC's Design Guide 21. The throat dimension refers to the shortest distance from the root to the face of the weld, which for a PJP weld is approximately the depth of the groove. (See Table J2.1 for actual effective throat since it depends on the welding process).

AISC lists allowable stress for weld (with Tension normal to weld axis) as 60% of Fexx, therefore they are still considering a shear failure for that type of loading.

The 0.6 factor on Fexx does not mean that a PJP weld with tension normal to the weld axis is actually a shear failure. See the commentary for Section J2.4 which says:

"The factor of 0.6 on FEXX for the tensile strength of PJP groove welds has been used since the early 1960s to compensate for factors such as the notch effect of the unfused area of the joint and uncertain quality in the root of the weld due to the difficulty in performing nondestructive evaluation. It does not imply that the tensile failure mode is by shear stress on the effective throat, as in fillet welds."​

You should be checking these welds using the strength reduction factors and effective areas provided in Table J2.5.

With respect to the CONNECTION, I'd have to disagree with your sketch for the Fillet tension parallel, bc yes the stress is P/A but for sizing fillet weld I would still take a failure plane at throat and do my weld calc based on that (each fillet resisting P/2).

For the fillet weld with tension parallel to weld axis case that I showed above, there is still shear in the fillet weld. You still have to check for the shear stress at the throat of the fillet weld. You just do not have to reduce the strength of the weld because it also has tension parallel to its axis.

 

[Wanna_be_SE]

That's interesting what you showed from AISC Design Guide 21, I'll have to read up on that. Thanks for sharing, I've never referenced that design guide before.

What do you mean when you say "You just do not have to reduce the strength of the weld because it also has tension parallel to its axis.", Table J2.5 lists shear as the only check and its using a strength reduction factor of 0.75 (LRFD), which is the smaller value (compared to 0.8 for tension on PJP welds). So what are you talking about when you say you don't have to reduce the strength, bc for fillet welds yes there can be tension in the weld itself, but we're talking about FAILURE, not stresses occurring prior to failure, and so the assumed failure stress would be a shear failure, so why would you consider anything else?

Please explain.

 

[ProgrammingPE]

You just do not have to reduce the strength of the weld because it also has tension parallel to its axis.

All I was saying here is that you can ignore that there happens to be some tensile stress in the weld when you check the weld for shear. The weld shear strength calculation (which includes the strength reduction factors) that you typically do is unaffected by the tension. The tension does not consume any of the available shear capacity like how some interaction equations work.

 

[Wanna_be_SE]

I was able to get a copy of AISC Design Guide 21, and I see the discussion on PJP throats.

I wanted to note that I think i understand why the Design Guide says to use the depth of groove as "throat", bc even if you took a shear plane at 45 degrees from toe of weld, you would actually have two shear planes - both legs of triangle created (hyp would be groove depth or "E" as defined above), and that's always greater than the groove depth or hyp, and therefore groove depth (hyp) controls.

Well this explains the tension in PJP welds, although I have to admit welds I typ see do not have a single bevel as illustrated above, I'd typ see a reinforcing fillet or something so seems like a somewhat "unique" condition if say you couldn't fit a fillet on the outside or something but nonetheless, good to know.

Thanks!