von Mises stress with algebraic sign 1

[mfinke]

Dear all,

for a part under triaxial stress condition i´d like to do a fatigue life estimation.

From a FE analysis i got the von Mises stresses (to compare uniaxial material props with triaxial stress in part)

material assumed to be ductile.
( so yield in tension and compression should be similar)

Question :

is it a legal way to add the alegebraic sign to the von Mises stress. I´d like to plot the results in a Haigh diagram where i take care of the algebraic sign of the mean stress. (just to visualize tension and compression)

 

[Drej]

Very illegal - you can get put in prison for less than that(!). To visualise tension and compression use the max and min principals.

 

[gurmeet2003]

mfinke,

Most of cases I deal with are proportional loading. The directions of principal stress do not change. I look at both the maximum and minimum principal stresses in three principal directions during the cycle and use the largest stress fluctuation for calculation of fatigue stress.

Thanks,
Gurmeet

 

[corus]

I'd plot the results of both the maximum and minimum principal stresses aa they vary. If you're showing the results to an audience that aren't familiar with stresses then rename them as tensile and compressive stresses, or pulling and pushing stresses if you're explaining them to higher management.
Your fatigue life calculation will of course depend on the principal stress range which you should calculate from the stress components if the principal direction varies. It has nothing to do with Von Mises stress which is for assessing stresses against a yield criteria.

corus

 

[feajob]

mfinke, Check the following link:

http://www.machinedesign.com/ASP/strArticleID/55730/strSite/MDSite/viewSelectedArticle.asp

AAY

http://www.geocities.com/fea_tek/

 

[Irwin]

I would suggest the next: sign(s3+s1)*Von-Mises stress.
You can use other equivalent stress with this sign function.
This is not good for everycase!!!! If the shear stress is high in this case this formula totally bad. This equation shows us that the volume of the "stress cubic" is growing or not. If it is growing in this case it is tension. The problem is the shear stress, because in this case the volume will not change, just the shape of it is deformed.
Other possible way is the "Critical cutting plane method", like in FEMFAT software. This is also not correct, but more than nothing.


Best regards,

Irwin