Weld type dependence of S - N (wohler) curves in fatigue analysis

[serdart]

Good Days,

A problem that I recently come across is confusing.

As far as I know, for a certain type of material, say 1040 steel, the slope of the wohler curve is constant (3 for normal stress & 5 for shear stress), for any type of weldment. But a document that we recently obtain tells that weld type ( fillet, butt...) and quality (weld finishing, grinding...) affects not only the endurance limits, but also the slope of S-N curve.

Does anyone have an idea about this slope issue?

Thx

Serdar tekin

 

[ishvaaag]

Code CTE seems to use different curves (of such same slope) for differently categorized details, then targeting your problem. You may want to give a look to Anejo C (Annex C) in

http://www.codigotecnico.org/index.php?id=33

download (Descargar) DB-SE A: Acero, that is the part that cares for the steel structural design and then go to Anejo C at page 147 and then to page 150 for the referred S-N figure.

 

[Ron]

Serdart...what is the document you reference? Welding, particularly quality issues, result in "jump" discontinuities in the S-N curve, not slope changes. Weld location and type could affect the slope of the S-N curve but I find it hard to believe that a quality issue would change the slope of the S-N curve. Weld discontinuities result in a reduced endurance, but it isn't a "smooth" transition.

 

[serdart]

Thx Ron, could you give me a referance material about:

"Weld location and type could affect the slope of the S-N curve"

 

[Ron]

The information is a combination of metallurgical properties and mechanical properties, information about which is gained through the study of both parameters. The American Society for Metals (ASM) has some excellent reference material on both.

 

[corus]

BS7608 gives the fatigue curves for different weld classifications. It's probably been renamed as a european standard these days. The same curves are given in the pressure vessel standard, BS5500.
The slope is generally 3 but the other constants used in the formulation differ for different weld types, locations, or principal stress direction. I'd refer to the appropriate standard.


corus