Tresca vs. von Mises 2

[julian89]

Good afternoon ladies and gentlemen,

Consider a case whereby a vessel has been designed in accordance with ASME VIII-1. At a later stage, due to local metal loss for example, the vessel is re-evaluated in accordance with ASME VIII-2 part 5 rules with an FEA. As we know, Div. 1 uses Tresca and Div. 2 allow the use of von Mises yield criterion. So, upon re-evaluation after Div. 2 rules, will it be permissible to use the von Mises yield criterion?


Here is a short article which touches this topic.

Julian

 

[TGS4]

It depends on what you mean by a later stage. If the local metal loss is due to operation, then you would follow the rules in API 579-1/ASME FFS-1. If the local metal loss is prior to operation, then the rules of VIII-1 should be followed, including the extent to which U-2(g) and Appendix 46 may be called.

The difference in stress invariant is mostly irrelevant to the discussion.

Btw, the article is incorrect. VIII-1 use max principal stress, not Tresca.

 

[Ehiman1]

I agree with TGS4, I have studied that you use Von Mises in ASME VIII div.I and Tresca in ASME VIII div.II!

 

[TGS4]

Tresca has not been used in Section VIII, Division 2 since 2006. However, many part of Part 4 still use Maximum Principal Stress. Part 5 of Division 2 solely uses von Mises as the invariant.

BTW - the other problem that I have with that article is that the failure mode of heads (especially torispherical heads - I detest the term F&D) is often knuckle buckling. That is not merely a stress "issue", but highly dependent of the geometry and the Young's Modulus.

 

[julian89]

Thank you for commenting!

I agree @TGS4. And using API 579 may lead us back to von Mises, correct?

However, what if 1) the client would like to know if the component is still in accordance with the original design code? Or 2)FEA calculations are required as part of a clients design requirements, in the design stage in addition to Div. 1 calcs (i.e. use Div 1 allowables via the methods of Div. 2)?

Thank you for your time.

 

[TGS4]

See Division 1, Appendix 46. Your answers are there. (Hint, applying Division 2 methodologies- including the choice of invariant - with Division 1 allowable stresses, is well accepted)

 

[r6155]

If actual thickness is greater than the required thickness no calculations are required.

Regards

 

[julian89]

Thank you for pointing me in the right direction!