Biaxial Yielding Failure Criteria

[mathlete7]

Hi,

I have a problem in which I have a metallic liner (AL6061-T4) in a pressure vessel which is designed to yield biaxially during pressurization. My question is, what would be the ultimate failure criteria for this condition? That is, what combined stress state would be the most appropriate to predict exceeding ultimate allowables?
I know that the von mises yield criterion is used to predict onset of yielding, however it loses meaning to use this criteria beyond the point of yielding from my understanding.
Would maximum principal stress/strain be appropriate to predict failure in a situation like this?

Thanks for your input...

 

[CoryPad]

The von Mises yield criterion is replaced with flow rules in the plastic regime. This allows estimation of stress and strain between yield and fracture. Max principal stress/strain can be used to estimate final fracture, as can fracture mechanics.

 

[Wrenchbender]

If you have a ductile material, with widely separated YS and UTS, a plasticity analysis would be the most thorough approach. Without resorting to finite element methods, I think you can do a decent closed form solution using a ‘Levy-Mises’ analysis. Check any plasticity book. I’ve seen a good treatment in Dieter – Mechanical Metallurgy. In the end, don’t stake your life on the answer because failure is a statistical phenomenon and the equations, including the VM or Tresca in the elastic range, are semi-empirical. To make things more complicated, you might also consider the presence of critical flaws (which cause failure at less than yield). A ‘leak before break’ condition is usually preferable and probably easily attainable with a ductile material such as yours.

 

[metengr]

mathlete7;
I was reading your post and am a little confused as to what your intentions are with respect to the vessel and liner. The pressure vessel would not be designed to yield if you are using any of the accepted pressure vessel construction codes. Most international codes that use design-by-rule do not allow for primary membrane stress to exceed yield in service based on use of allowable stress values.

If your discussion pertains to the liner itself, the criterion for failure will be allowing for some acceptable level of permanent distortion if you are assuming that the liner will be subjected to service stresses that exceed current code allowable stresses. You will also need to evaluate what damage mechanism prevails in service. If this pressure vessel is cycled or you have thermal cycling, low cycle fatigue will be the controlling variable for this liner.

 

[Compositepro]

What kind of failure are you trying to analyze for? My understanding is that liners typically fail due to compressive buckling when the vessel is depressurized. The liner can't fail before the composite under pressure.