Stress criteria 1

[ankser]

Dear members,

I doing a static structural nonlinear analysis (Bilinear material curve with contact pairs) of a steel structure using ANSYS classic.

The von-mises stress is less than yield strength but the first principal stress exceeds the yield strength (both von-mises & first principal stresses are developed at same location).

How the result can be concluded, based on von-mises or first principal stress?

Thanks in advance for your supports.
Ankser.

 

[rb1957]

this is, in my mind, the problem (and the strength) of von mises. vm combnes multi-axial stresses to give you a uni-axial equivalent stress (to compare with standard material allowables). multi-axial tension will produce the situation you describe.

the question you need to ask yourself is "is the FEM prediction really valid ?" have you got a solid model with 1 element through the thickess, and so unreliable stresses ? is there something in the loading that might reduce the beneficial multi-axial tension ??

you can use either vm or max principal as a failure criteria; don't IMHO use the minimum of both, that's an each way bet ... it's conservative to be sure. vm is a perfectly valid failure criteria.

 

[corus]

If you have plasticity then your Von Mises stresses are bound to be less than yield.

Do an elastic analysis and assess your stresses using design code standards for your particular structure.

Tata

 

[ankser]

Thanks rb1957 & Corus.

rb1957, I am using SOLID92 (10 node tetrahedral) elements with good density across the thicknesses.
Also, yes, there is muti-axial tension due to loading.

Thanks,
Ankser.

 

[rb1957]

it sounds like the model is reasonable, tet10s are about as good as you're going to get, several through the thickness sounds good too.

you might calc a vm stress based on the two in-plane principal stresses.

 

[Elabbasi]

The von Mises stress and the maximum principal stress are two scalar values obtained from the stress tensor. You can easily come up with a stress state for example that has zero von Mises stress and a large maximum principal stress. Which stress measure is more important depends on the material. Failure and plasticity in metals for example is dependent on the Mises stress.


Nagi Elabbasi

http://www.veryst.com