von-Mises Stress vs. AISC allowables 1

[monty01]

I've always considered von-Mises stress results as an "in-depth" analysis and not just a basic average stress across a section.

My question is this: Is it appropriate to compare a von-Mises stress against AISC allowables. Previously i've compared it to yield on the basis that it isn't an average stress result. I believe that the AISC allowables are used when average stresses are calculated/approximated.

Does this seem correct? More to it than this? Any help would be much appreciated.

 

[UcfSE]

Is your analysis capturing instability, not just yielding? Many of the AISC equations take into account instability, such as lateral-torsional buckling or local buckling. If your analysis and resulting von Mises results do not account for this, then I would say no you cannot compare that to AISC results. Otherwise, or if instability is no problem in your case, then you'll need to put a safety factor against your results. Depending on your application, that safety factor may come from the AISC, or some other governing code or body if you're something covered by AISC, such as a machine part. It all depends on what your code allows.

 

[pylandjw]

AISC values are for principle stresses according to basic beam calcs. For tension, you are allowed .6Fy and for shear you are allowed .4Fy.

But von Mises represents a combined state of both shear and tensile stress.

One way to compare is to substitute the AISC limits into the von Mises equation to get:

((.6*Fy)^2+3*(.4*Fy)^2)^(1/2)=.9Fy

However, codes we work with, if they specify that von Mises stress can be used, only allow von Mises stress to go to 2/3 of yield.

 

[johnhors]

Principal not principle !

 

[monty01]

No instabilities in this analysis as it is a forged pipeline flange. It's more a matter of how i've noticed some in my group use von-Mises as though it was a tensile stress. Thus i've seen some unity checks in the form of UC= vm stress/.6yield, which have always seemed over-conservative to me as von-Mises is NOT a tensile stress, but a combination of principal and maximum shear stresses.

Thanks pylandjw, i would have to agree on your assessment. Although i haven't seen much in AISC code covering von-Mises analysis. If you have, could you point me to the section it is covered?

 

[jte]

No instabilities in this analysis as it is a forged pipeline flange.

If the component being evaluated is a pipeline flange, why use a structural code to provide an acceptance criteria? I've never seen a flange fail due to being overstressed. If anyone out there has a photo of a piping flange which has torn open or buckled, I'd sure like to see it. The "failed" flanges I've dealt with have exclusively failed through leakage at the gasket surface - which is a strain/deflection failure (or gasket or installation…), not a stress failure per se. I generally deal with ASME codes and thus, for a flange design I would usually default to an ASME VIII Div. 1 Appendix 2 approach… If going FEA for some reason, then Div. 2 App 4.

jt