Double fillet weld corner joint

[weh]

We manufacture steel connections for the heavy timber industry (glulam). Our typical weld detail is a fillet weld as shown in the attached drawing. I recently had an Engineer say that this was not an approved weld for the given loading (top mount hanger). does this joint not fit under prequalified welded joints - fillet welds Table 8-2 AISC Steel Construction Manual?

 

[gtaw]

You should limit posting your question once.

Best regards - Al

 

[TLHS]

Regardless of whether it's prequalified, it's not a great weld configuration. Setting your plates in position is going to be a giant pain. Also, your load path for moment resistance seems kind of wacky if there's any moment transfer (I assume their must be some given the plate arrangements).

I'm not a weld engineer or anything, but I suspect that regardless of whether it meets the letter of the prequalification it may not meet the spirit of it, given the way the roots basically meet like that. You're also going to fuse the same base metal twice because they're so close.

 

[TLHS]

The guys in the other thread clearly know more than me! I still don't like the looks of it, but that doesn't necessarily mean it's wrong.

http://eng-tips.com/viewthread.cfm?qid=358309

 

[dhengr]

Weh:
For your static application, that weld detail should probably be o.k., but there are several issues that you have to resolve in your own mind. I’m not a great one for worrying that some code approves every exact detail, in every respect. I would sooner that we understand the basics, the areas of potential troubles, and then use good engineering judgement, design and engineering mechanics. Gtaw gives good counsel, and I am not sure that you need a new WPS for every possible detail. But, you do need to know welding design and your own processes and welders, so you know what they can and can’t do.

On the detail you show, the welds can certainly be sized to take the static loads. But, the potential problem that I see, and the objection that the other engineer may have had, is that the flow of the tensile stresses/force fields is perpendicular to and flows across the two unfused roots of the two fillet welds. And, this is not a very acceptable condition, since it can start cracking at the roots. There is a good chance that the first weld will penetrate through the small root face (land) which you have; and the second weld will remelt this, boil-out any impurities and give you essentially a full pen. weld. You should run some samples of this, and test your welders to prove that this is true. Alternatively, you could grind or gauge that root area before making the second weld, then there should be no question of trapped junk causing imperfections which are perpendicular to the stress field. You also have to be careful how you start and stop these welds at the two edges of the plates, so that these areas are properly finished and don’t become crack starters. You could likely break form that top flange on the back plate and hide any minor length variations behind to bottom bearing plate.

 

[dcarr82775]

IMO you don't really have real moment continuity at the joint, which you need

 

[a2mfk]

Agree with Dcarr and TLHS. What happens when this connection sees a little bit of axial load along the beam axis, and then you have a moment at that weld?

Why not use a piece of bent plate and/or angles to build this connection?

Sort of like this:

http://www.strongtie.com/products/connectors/GLS-HGLS-GLT-HGLT_GLULAM.asp?source=glulamcat#

 

[gtaw]

A sketch of the actual connector and the design loads might clarify the question.

Best regards - Al

 

[racookpe1978]

It is important to note that the "corner" details of ship construction for frames-intersecting-hull-plates, and the ribs-intersecting-frames, and for tank-walls-intersecting-internal-tank-ribs (where watertight integrity or oil-tight integrity are not important, ALL "cut-out" or use small "arched" wall-to-wall connections. This is because of fatigue concerns (no residual stresses in the tight 90-degree angle corners, faster welding (no fighting against access down in the corner, no weld-over-weld concerns as each bead intersects earlier beads, no grinding or deep-pocket end-grinding and "impossible" corner cleaning.

Unless this is a fluid-tight corner, don't weld it like you propose.

 

[weh]

gtaw attached is a sketch which I hope clarifies the situation. This is a typical timber beam to timber beam connection. The connections are all custom made for specific projects. The connections usually consist of different plate thickness,skews,slopes, etc so bent plates are usually not an option. The plates making up the connections are laser cut to shape and assembled using open corner fillet welds as shown in the sketch (using a slight overlap = 1/16"). Although it requires more weld material than a complete penetration groove weld, there is no joint bevels required.

 

[a2mfk]

Still not sure why a bent plate would not work for the bearing portion of the connection (the main "L"), and then you can weld shear tabs or a built-up bucket type connection to that. Think- what if there is a small amount of moment transferred into that welded connection you are proposing, which could come from a small amount of axial tension due to member shrinkage, loading, other structural movements, etc.?

 

[gtaw]

I see from the sketch that the hangers are not designed to transfer moment and the loads are static. That being the case, the welds are not subject to moment or fatigue. The llet weld is designed to be loaded in shear, an acceptable condition per AWS D1.1.

The structural code has no limitations on the size of the fillet weld unless the joint is a lap joint ref. AWS D1.1-2010, clause 2.4.2.9(2)). The weld can be built out to the full thickness of the plate in an outside corner joint and it can be as large as necessary on the inside o the corner joint.

The double side welded joint is a partial joint penetration weld unless there is a back gouge operation. As PJP or a fillet weld, the efficiency of the joint is limited. The maximum unit stress is 0.3 time the tensile strength of the filler metal, but not more than 0.4 times the yield strength of the base metal. AWS D1.1-2010, Table 2.3 provides the designer with the allowable stresses per the structural welding code.

In any case, the details of the design are subject to the review and approval of the Engineer representing the Owner's interest. As a fabricator, I would suggest the standard designs/details be reviewed and approved by a registered structural engineer.

Best regards - Al

 

[ztengguy]

wouldnt a bent plate be cheaper, no welder to pay, plus you get the moment continuity.

 

[tony1851]

+1; cannot understand the logic of the double weld.

 

[gtaw]

Look at the edge of a wooden beam or a laminated beam. They are is pretty square, not rounded like a bent plate.

I can see where the fabricated hanger would conform to the beam better than a bent plate.

Best regards - Al

 

[weh]

Thank you for your responses, especially gtaw. The main reasons bent plates are not an option are because these connections are typically sloped, skewed, consist of plates of differnet thicknesses.