How do I calculate the stress on a lap fillet weld in bending?

[Fantastic Fox]

I have to design a lap fillet weld joining a plate to the top flange of an beam. There will be a force acting down on the plate so the weld will be bending. How do I calculate the stress on the weld?

l l <- Forces acting on the plate and beam
----------------- The plate is the length of the beam

 

[PhanthomJay]

For cover plate weld design, you have to calculate the longitudinal shear at the intersection of the cover plate and top flange of the beam. Does tau=VQ/It sound familiar?

 

[Fantastic Fox]

I haven't used that equation before, so I have a few questions:

Is t = thickness of the flange or the breadth of the flange? Or, is it the thickness of the flange or the breadth of the plate?

When I have calculated tau, where do I go after that?
Force per inch of weld direct vertical shear = f = P/A? Then,calculate Force per inch of weld due to bending too? Add them, and divide by allowable force per inch of leg?
tau = shear load / (throat thickness)(length) to calculate t?
Thanks

 

[PhanthomJay]

yow, it's been 20 years since i designed my last weld. Too much managing and scheduling, and too little engineering. My recall is that you want shear flow, VQ/I , at the plate/beam interface (VQ/Ib gives you the shear stress in the beam at this interface, where b is the width of the flange, not its thickness, but this is irrelevant for the weld design.) So calculate 'I' for the built up member about its neutral axis, and Q (the moment area) of the plate (Q=area of cover plate times the distance from its centrioid to the neutral axis), and your result will be in kips per inch; usually you will have 2 welds on either side of the plate, so divide the result by 2, and that's the load per weld, and size its thickness accordingly (usually it's a minmum size fillet weld, because shear stress is not that big near the top of the section). BTW, the weld resists the bending load in shear, so this is weld shear stress, not bending stress.

EDIT: i just noted your query about adding the shear flow stress to the vertical shear V/A stress. The answer is NO. V/A shear stress is the average of the shear stress across the entire section. You don't want the average, you want the actual VQ/I at the interface. So just calculate it, and size the weld to take that force/in, and make its length continuous over the full length of the cover plate. If the weld thickness comes out less than the minimum required by codes, it will be oversized, so instead of making it continuous, you can make the weld intermittent at so many inches on center, if you want.