Weld / Base Metal Stress Concentration?

[mojo41]

How do I evaluate the nominal base metal stress at the toe of a fillet weld? The nominal stress is 19.2ksi and less than the plate's yield strength (30ksi), however I know a localized stress concentration (i.e. weld toe) exists at this location. I have searched a number of codes (AWS, IIW, AISC, etc.) and I can't find any mention of stress concentration under static loading conditions. Are these stress concentrations to be ignored, or is it cooked into the code?

For AWS D1.1, should I convert the nominal stress (Von Mises) to maximum shear stress and compare that to 0.4 * yield (12ksi)?

Thanks

 

[metengr]

http://www.scielo.br/pdf/si/v22n2/0104-9224-si-22-2-194.pdf

 

[TimSchrader2]

Hello. It looks like the torsion from the vertical load is somewhat being transfered to a stiffener next to the 30mm thick ear. As the tube thickness is almost 1/3 of 30mm the stiffener will help. By stress concentration I assume you are talking about the local loading in the base metal on the thinner tube and possibly a stiffener. So the tension created at this area will mostly go thru the stiffener. Simliar loads/stresses are at the top of the ear, so a stiffener may be needed there also.

FEA would help confirm.

 

[mojo41]

Thanks for the input. However, I don't actually want to know what the SCF (stress concentration factor) is at the toe of a fillet weld. The real question has to do with how codes deal with the SCF at the toe of a weld and how that is used to evaluate the base metal under static conditions.

Also, I don't need to answer the example presented and it is only being shown to illustrate the problem

 

[dhengr]

Mojo41:
You may actually be confusing two welding terms, fillet ‘toes’ run the length of the weld where the sloped face of the weld meets the base metal, and they may leave an undercut, which is a stress raiser, under many conditions. Your worst stress concentration is actually at the fillet and groove weld ‘termination’ where your arrow from “Von Mises Stress (Base Metal),” your sketch, points to the termination of those welds. There you have a combined stress condition which is the tension from the canti. moment (352kN)(128mm) and the shear stress/shear flow (352kN)/(260mm, weld length). You should stop those welds short of the end/corner of the lifting lug (250-260mm, not 283.8mm long), and not weld around the corners of the lug. Those corner welds are almost impossible to do well, without leaving a notch/stress raiser at the corners of the lugs. But, you have an additional adverse condition at that same location, and that is the plate bending (prying on the weld) in the tube side/web wall perpendicular to the plane of the tube wall.

That’s really an awful lot of weld, and weld prep. for the grooves on the lug for those welds. I think I would try to do that with fillet welds only. A 1/4” or 5/16” fillet both sides would probably about do it. If you had the option, on that detail, I would extend the lug up a couple inches and down about 3” and extend these legs out to the left 3 or 4” to the left, so they were bearing on the t&b flgs./walls on the tube. This would allow me to get about 3” of weld to the t&b flgs., getting that load reaction into the t&b flgs. where I want it, and eliminate the wall/pl. bending mentioned above. I would also make that lifting lug pl. 5/8” or 3/4” thk. and then thicken the doubler doughnuts to make up the diff. and account for the pin bearing problem.

There are plenty of examples in AISC Std. details and regular fab. details which show that under normal conditions, when the welds are done properly, and well, that the combined stresses or the small stress concentrations at weld toes or weld terminations can usually be ignored. Under normal conditions they have just not proven to be a problem. End shear connections btwn. perpendicular beams or beams to cols. are one example of a condition similar to yours. That does not preclude the need for attention to these conditions under some conditions, that’s called ‘engineering judgement and experience’ or lack thereof when it causes a problem.

You’ll have to show us how you calc’ed. the Von Mises stress by hand, that’s tricky. Von Mises stresses are not nominal stress usually calc’ed. by hand, are they? I would calc. the shear flow btwn. the lifting lug and the tube, and the canti. moment bending stresses, making some assumptions about how that bending stress is distributed over the height of the lug, and then make some more assumptions, engineering judgements, about combining those stresses.