Minimum Number of Elements Through Thickness for Stress Linearisation 1

[Mohit_singh]

Hi ALl,
I am doing FEA of a Nozzle using ASME VIII div 2 . Now I know from someone that we have to take minimum three elements along the thickness for Proper Analysis. Can someone give me ASME reference to the same ?

 

[TGS4]

There is nothing in the ASME Code that will replace your engineering judgement. 3 elements in the through-thickness direction would be a minimum IF you were using quadratic elements. 3 linear elements through-thickness would be completely unacceptable, however.

The matter is that you need to adequately define the through-thickness stress distribution, including both the linear and non-linear portions - especially if your pseudo-elastic stresses exceed yield.

Much more discussion on this topic can be found in ASME Section VIII, Division 2, Part 5, Annex 5-A.

Also, please remember that you must meet the requirements of ALL of Part 5, which includes Paragraphs 5.2, 5.3, 5.4, 5.5, and particularly for nozzles, section 5.6.

 

[Mohit_singh]

Thanks TGS4 for your answer. However if I am using linear elements than how many elements are acceptable?

 

[TGS4]

You're the engineer - you tell me.

 

[Mohit_singh]

HUH!! I use more than three elements as the WRC 429 and ASME PTB uses in its examples.

 

[TGS4]

First, you need to ensure that your solution is mesh independent. In general, I don't think that three linear elements through-thickness is adequate. Depending on the situation, 10 or more may be needed. There are no rules, just engineering judgement - and your judgement needs to withstand severe scrutiny.

Second, I sure hope that your linear elements are not tetrahedrals.

Third, the examples in PTB-1 all use quadratic elements.