Weld capacity??? 1

[PowersPE80]

How would you design the weld in the attached detail? The moment is very small (0.5 ft-kips) but I would like to have a calc to back it up.

Thanks.

 

[sbisteel]

I'd convert the moment into an equivalent force x distance, and use AISC's eccentrically loaded weld group tables (table 8-4).

 

[haynewp]

Find the section modulus of the 2 weld lines added together of whatever thickness welds you choose (x 0.707), then use M/Sweld to find the max tension in the welds.

 

[Archie264]

As an aside, you just about need bionic eyes to read the diagram in AISC's Table 8-4. It's a reminder that "classical" drafting is becoming a lost art in this age of computer modeling.

 

[dhengr]

PowersPE80 :
As second and third ‘asides,’ that’s an awful detail and the angle extension will cause the whole beam be loaded torsionally. The angle’s carrying cap’y. is biased toward its vert. leg, so it will tend to twist the beam, or else its shear center should be centered over the web of the W4. I would pick a 3" S shape for the extension, and center it on the W4. Then, your welds are out near the flg. tips, so flg. bending, strain and deflection should be considered, as they relate to prying at the roots of your longitudinal fillet welds. Finally the end terminations of your welds are particularly sensitive to stress concentrations by the nature of the weld process, and your arrangement essentially has the potential of infinite weld load or stresses at these weld terminations. They will just unzip from one end or the other. You might increase the lap length to something greater than 6" to improve this situation, but in its simplest form your moment is resisted by a weld load couple with a lever arm of about 5". I would increase the lap length, and put cross welds from flg. to flg. at both ends. These welds would/could only be about 2" long, but they would be centered over the webs of the two members and be much less susceptible to high stresses (unzipping stresses) at weld end terminations. I would leave the light longitudinal welds, just terminate all the welds short of an edge or corner. Do not weld all around, and over a corner, or you just create another notch.

 

[TLHS]

I wouldn't connect an angle like that generally. The only way I'd be comfortable connecting an angle in moment is if I could connect both sets of legs. As it stands, your centre of resistance is going to be in an odd place because your toe is significantly more flexible than your heel.

If I *really* wanted to do it for a small load and it wasn't worth wasting a bunch of effort on the math, I'd likely design the system so the whole moment was resisted by the heel weld. Then I'd take that system and resolve the torsion by putting together a couple between the heel weld and the toe weld. I'd then check the angle leg and make sure it could actually hold up in moment to transfer the shear that's resolving that torsion. Then I'd add the welds necessary for those two purposes at the heel and use that weld on the heel and toe.

It's not the load path that's actually going to happen, but it's a conservative load path and it's reasonable to model without doing FEA or a complicated stiffness analysis.

 

[TLHS]

Of course, if it made an unreasonable weld, I'd look closer.

I'd also at least pull the welds around the corners a bit, even if I didn't design based on it

 

[haynewp]

Some discussion on returning welds for statically loaded structures:

http://www.modernsteel.com/uploads/issues/may_2008/052008_30777_steelwise_web.pdf