Allowable Limit on Primary Local + Primary Bending Stress; Sec. VIII, Div. 2, Part 5 1

[m_ridzon]

For context, I'm looking at 2017 ASME Sec VIII, Div. 2, Part 5. I'm questioning the P_L + P_b allowable. Part 5.2.2.4 Eqn. 5.4 states that the limit is S_PL, which is 1.5*S. However, Figure 5.1 states for primary bending, "excludes discontinuities..." Secondary bending states "occurs at discontinuities..." So presumably, primary bending at/near a discontinuity, is Q stress. The allowable for P_L + Q is 3*S. So for the P_L + P_b allowable, I cannot tell if the allowable should be 1.5*S or 3*S. By definition, P_L is the membrane at/near discontinuities. But P_b excludes discontinuities per Figure 5.1.

 

[TGS4]

First things first, please always use the current Edition of Part. Changes are constantly being made.

You have correctly stated that the allowable limit for P_L + P_b is S_PL. However, you are incorrect to state that S_PL is equal to 1.5*S. S_PL is actually the greater of 1.5*S or S_y (with some caveats on the side).

You are further correct that Figure 5.1 notes that primary bending excludes discontinuities. Therefore, bending at discontinuities is typically secondary (with the caveats as noted in Table 5.6).

Therefore, there is no such thing as primary bending at a discontinuity. It is simply bending, and then you as the engineer need to classify the stress (primary or secondary) which then leads to a categorization (P_b or Q). In your case, bending at a discontinuity (depending whether it is in a cylindrical or conical shell attached to a head, or it is in a flat plate connected to a cylindrical or conical shell) would be classified as primary or secondary, and then assigned a category of either P_b or Q.

Of course, this is all in the context of applying a design load combination in accordance with Table 5.3. Again, this is only for the failure mode of Plastic Collapse. There is no limit on P + Q for Protection Against Plastic Collapse.

For the failure mode of Ratcheting, you have a different set of Operating Load Combinations (see, this is why you need to be using the current Edition - the changes that I am going to talk about here were made in the 2019 Edition). Then, the limit on the range of P + Q is S_PS. BTW, the limit S_PS is the greater of 3 times the average value of S over the cycle in question (which may varying depending on the temperature of the individual cycle being analyzed) or 2 times the average value of S_y (with the same note as the average value of the allowable stress).

So, it not merely a matter of determining whether or not a limit is 1.5*S or 3*S. Even asking about such a choice is incorrect. Focus on the specific Failure Mode, first. Start with Protection Against Plastic Collapse. Apply the Design Load Combination(s). Then, after having separated the loads as noted in 5.2.2.4, especially Step 1 and Step 3(a). The make your stress classification judgement of whether the bending stress is primary or secondary - taking guidance from Table 5.6. Once you have judged the stress classification, you will determine the categorization. And once you have the categorization, then the Code will clearly tell you what the allowable is.

Finally, I highly recommend reading and reflecting on paragraph 5.2.1.2.

 

[m_ridzon]

@TGS4, thank you very much for this insightful feedback. I think I know what I need to do now.

In terms of using 1.5*S vs. S_y (and also 3*S vs. 2*S_y), I'm aware of that distinction. I didn't mention it in my post because I didn't want to unnecessarily add detail to confuse the main point of my question. Thank you for the clarity.

In terms of using the latest version...in a perfect world, you are probably correct. However, the version to be used for a consulting project is typically dictated by the customer and/or municipality where the equipment will reside. In the case of a refurb and/or requalification (e.g., nuclear power plant equipment with extended life), I have frequently been asked to use the version that was used for the original design decades ago. ~shoulder shrug~ So I don't always have the luxury of using the latest-n-greatest. But otherwise, I'd agree with you.