Fatigue Design/welds

[ToadJones]

In short, is it fair to say that designing welds for fatigue involves first designing the weld for static loading, then ensuring that the weld stress is below the fatigue design stress range?

 

[rowingengineer]

It is one way to skin the cat.

A clear conscience is usually the sign of a bad memory

 

[ToadJones]

Row-
Is that a yes, kinda?

 

[dhengr]

Toad:
Whatever the fatigue design stress range is? It will most always be less than the allowable static stress in the weld. It is also fair to say that a weld line parallel to a stress field will always fair better than a weld line across the stress field, because of the potential stress raisers the latter introduces across the stress field. Good clean design and detailing to keep the stresses flowing smoothly is all important in minimizing stress concentrations, which invariably widen the stress range. No abrupt section changes, no reentrant corners, no weld imperfections, etc. etc.

 

[ToadJones]

dhengr-
I guess that is why I am asking....I think I may be brain- farting here, but, say in the case of a welded plate girder- one would think the design of the bottom flange-to-web weld would only involve the static load case (even in designing for fatigue) unless there was a stress reversal, no?

 

[dhengr]

The weld btwn. the flange and web is parallel to the primary stresses, those of most interest from the fatigue standpoint. And, the stresses in the weld are also parallel to the primary stresses. Its static strength is higher than, but fairly near, its fatigue strength, and generally as good as the base metal if you are using the right electrodes. Even weld defects like undercutting and the like are generally parallel. A properly made and tested CJP weld in the flange is generally as good as the base metal, but defects are likely to be across the stress field, thus the NDT. And the weld is best ground down to match the base metal thickness and to eliminate undercuts, etc. A change in flange thickness or width required the bigger flange to be faired down to match the smaller flange. A cover plate with continuous side fillets is better than one with skip welding, because each start and stop is a potential stress raiser. A fillet weld across the end of the cover plate causes all kinds of problems; change in section, potential flaws, undercuts, etc., and stresses trying to flow around the root, an awful fatigue condition.