Maximum Economic Size of Fillet Welds Using SMA Process

[NoWittyHandle]

What are the economics of using different size fillet welds in steel using shielded metal arc (stick) welding? I am designing the wales (steel beams placed against the excavation wall to support the wall) for an excavation, and the specifications require us to develop the full moment and shear capacity of the beam at splices. The contractor doesn't want to use groove welds because the welds would have to be x-rayed.

I'm guessing that it is usually cheaper to use longer lengths of smaller fillet welds than to use shorter lengths of larger fillet welds. For example, a 3/8" fillet weld might take weld passes, but a 3/4" fillet weld might require seven or eight passes. I don't know if it is even pracitical to use a 3/4" fillet weld, and in the past I have usually kept fillet weld sizes to under 3/8", but this application is going to require tons of welding. Thanks for your responses.

 

[metengr]

EricinNJ;

I am designing the wales (steel beams placed against the excavation wall to support the wall) for an excavation, and the specifications require us to develop the full moment and shear capacity of the beam at splices. The contractor doesn't want to use groove welds because the welds would have to be x-rayed.

Please re-read the above carefully. Explain what "specifications" means. Just because the contractor does not want to x-ray welds should not be a reason to deviate from the original design. I see some serious problems here.

 

[HgTX]

Yes, longer smaller welds are more economical than shorter larger welds.

Blodgett gives a rule of thumb that fillet welds are more economical than groove welds if the required weld throat is 3/4" or less. But if your design calls for complete joint penetration, you can't just substitute a pair of fillet welds without consulting with the designer.

Hg

Eng-Tips policies: faq731-376

 

[GRoberts]

Nonwithstanding the groove vs fillet issue, on fillet welds, it is all about weld volume. Whatever fillet weld that gives the requried strength for the least volume of weld will be the most economical. With fillets, that ends up being smaller continuous welds.

Picture a small and large fillet. The failure mode is throgh the throat of the fillet weld. The metal that is in the fillet weld that does not contribute to the throat is just extra material. It is requried to form the fillet, but does not add to the alowable strength. So in a small continuous fillet, the metal surrounding the throat is much smaller than on a large intermittant fillet weld.

 

[MintJulep]

The spec seems incomplete, unless you are not telling us the whole thing.

Do you have to keep the stress under some limit? Are there fatigue considerations?

 

[metengr]

Getting back to the original post, IMO fillet welds should not substitute for full penetration welds by design, period. Why, because fillet welds are natural stress risers (inherent stress risers) that result in significantly reduced fatigue strength and limit load-carrying capacity. They serve a limited purpose in design and cannot be used without some strength reduction factor.

 

[metengr]

I should also mention that if RT (x-ray) is not possible or is too much of a problem, I would consider ultrasonic examination (UT) for acceptance.

 

[NoWittyHandle]

The specifications do not call for a particular weld type, so there is no inherent problem in using fillet welds. The beam would be spliced by placing a cover plate on top of the flanges. The cover plates would be made wider or narrower than the flange, providing a shelf to make the fillet weld. The stress limits on fillet welds are specified by the American Institute of Steel Construction and the AWS, and fatigue isn't a problem.

I spoke to a welding engineer who gave me estimates of the number of passes required to make different size fillet welds. He said that a 1/4" fillet weld requires one to two passes, a 3/8" weld requires four to five passes, and a 1/2" weld requires about ten passes. He added that fillet welds bigger than 1/2" are uneconomical.

Since the specifications require welds to develop the full moment capacity of the beam, shifting the splice to a location with lower loads isn't an option. As I suspected, while using fillet welds to make the splices is practical from a design and safety standpoint, it will be hugely expensive.

 

[HgTX]

Those numbers of passes seem high.

Hg

Eng-Tips policies: faq731-376

 

[unclesyd]

I'll have to agree with HgTX about the number of passes. I would ask the engineer to qualify his estimate.

Also ask about increasing the size of the electrodes if not restricted by the weld procedure.

 

[GRoberts]

Of course if you are really worried about economics, why are you using the SMAW process. SAW, FCAW, or GMAW would all be much more economical and the weld quality can be just as good.

 

[HgTX]

To get more specific, 3/8" really ought to be 2 passes, or 3 for those who just don't like the look of a 2-pass weld. You can extrapolate from there for your 1/2" weld. When you ask your welding engineer where on earth they got their pass estimates from, do let us know.

I also agree with GRoberts about choice of process.

Hg

Eng-Tips policies: faq731-376