Yielding of welds 1

[BMart006]

I completed a calculation package a few weeks ago regarding the modification of some lifting lugs welded to a rigid frame. The weld strength calculated from the ultimate strength of the weld metal (70 ksi min.) is adequate for the load with a >5 FOS, but one of the comments I got back was "what is the weld's FOS on yielding?" The yielding of the joined materials has already been addressed elsewhere in the calcs, so my question is: do welds even yield? and if they do, how do I calculate that condition? I have yet to see it addressed in AISC or ASME BTH. Thanks in advance for your help.

 

[SwinnyGG]

Welds certainly can yield. If you're specifying an E70 filler (70ksi min. tensile) that weld will have a yield strength around 58ksi.

To me this question would only become a problem if you're welding material with a higher yield strength than the filler- but you (should) never be doing that. If the weld itself is stressed beyond yield but below UTS, the base material adjacent to the weld is stressed to the same level, and will already be in the yield zone.

 

[BMart006]

If I'm understanding correctly, the stresses in the base metal and weld would only be similar at their interface (leg), but for the fillet weld (which is what I have), the weld stress is calculated across the 45-deg. effective throat. So if I were to try and check against the weld's yield strength, shouldn't I be doing this along that throat and not the leg? I also am not quite clear on why AISC does not require checking the base metal at the leg, but rather across the thickness of the part. I would think that if the leg is less than the part thickness (as usual), this would be the controlling mode.

 

[MGZmechanical]

I'm not used to ASCI but in Europe the beam section is checked against the yield strength and the weldments are checked against the weld throat with the ultimate strength of the base material. You do not consider the properties of the filler metal (they must be better than the base)

 

[SwinnyGG]

With a fillet weld, the leg length is not what's important. The throat depth is the effective part thickness through which any stress is distributed.

The throat thickness can also be deeper than the simple geometric throat of the weld, and possibly thicker than the leg length, parts being joined, or both, if deep penetration fillets are used. See attached image for clarity.

If the thicknesses of the parts being joined and the weld are imagined as tubes, and stress flowing through them is imagined as water, the diameter of the weld 'tube' is the throat depth- NOT the leg length.

Another way to picture this is that if you were to slice the weld through its neutral axis, and then measure the cross section, you would be measuring weld length and throat depth.

Ultimately, it sounds like you already have a weld stress level from previous calculation. If you have this number, compare it to the weld metal yield stress at the specified weld size, and you have a safety factor for weld yielding.

 

[dhengr]

BMart006:
Fillet welds normally act and fail in a shear (shear stress) mode, at the least weld failure area region, which is the throat of the fillet, irrespective of their exact loading orientation. You should pay some attention to the weld stress at the leg dimension too, particularly when you are welding to a lesser yield stress material. The yield stress in shear is approx. (F_y)/(3)^.5 = .58(F_y) for steel, where F_y is the yield stress in tension. And, your allowable weld stress is based on this shear yielding stress. Usually, you don’t have to worry about the shear stress at the leg failure area because the leg is longer than the throat and the filler metal and base metal are matched, but at the leg the weld puddle is mixing the weld filler metal and the lower strength base metal, in exactly what proportions, you don’t know. So, the shear stress should be based on the lower base metal shear strength. At the throat, you can be reasonably confident that the allowable weld shear stress is that of the filler metal. With your problem, LRFD and ASD the way they have it screwed around today, to match LRFD thinking, there meaning and results are kinda suspect. At least you really have to understand all the implications of this design thought process. Load factors, resistance factors, etc. are in place to keep yielding within limits under normal loading, service loads. In your case, you have a failure mechanism once there is any significant yielding, because there is no redundancy in the system or at the/your detail. In your case, significant yielding means strain and extension without much added load, whereas in the normal structural situations/details you can tolerate some yielding (start of hinge formation, etc.) because some distribution and added load cap’y. exists before a failure mechanism forms. Your details and design are much more dependant upon nice clean load paths and quality welding and detailing so as to minimize/eliminate any weld defects or stress raisers which might lead to sudden crack initiation/failure. If my details and welding are good and clean, I will tolerate a little confined yielding, knowing that there will be some internal stress redistribution and the stress/strain can’t really go anyplace, it is confined/controlled by surrounding material volume. You’ll see these high stress areas, reentrant corners, triaxial stress regions, stress raisers, etc. if you try to analyze your details with a FEA program.

 

[Tmoose]

I think on thicker weldments, that are not PWHT for stress relief, the as-deposited welds have yielded considerably during cooling, and upon final cooling much of the weld has residual locked-in tensile stresses approximately equal to the yield strength of the metal.

http://www.lincolnelectric.com/en-us/support/welding-how-to/Pages/weld-distortion-detail.aspx

https://app.aws.org/wj/supplement/WJ_1997_06_s233.pdf