Distinction Between Low and High Allowable Stress Values For Austenitic Stainless Steels

[RPRad]

This probably has been posted before.....I note the Code doesnt really doesnt provide much guidance with regard to when you use the high or the low allowable stress values for 316 and 316L SS, aside from advising you to use your best engineering judgement and to consider the lower values when designing flanged joints etc. Generally I have just used the temperature of operation and / or the magnitude of the design pressure as a basis.

Anybody got any a more defined criteria or useful words of wisdom?

 

[Ripz]

Where are you getting your stress values? I'm looking at the Section 2, Part D table, there's only one stress value for the applicable temperature for your ASTM Standard.

 

[marty007]

My general rule of thumb is that I use the lower stress values if deformation could lead to leakage, and the higher stress value everywhere else.

Using this basis, flanged connections and shaft-seals become really the only components where I use the lower stress values (these are just off-hand examples, I'm sure there are many more).

Pressure and temperature don't really affect my choice.

 

[TGS4]

What's really important in this discussion is understanding the (potential) failure modes, and how the different stress levels play into them.

As had been said, the lower values are typically used for deformation-sensitive components (flanges, etc). However, you need to have an understand of why that is.

From Appendix 1:

Two sets of allowable stress values are provided in Tables 1A and 1B for austenitic materials and specific nonferrous alloys. The higher alternative allowable stresses are identified by a footnote to the tables. These stresses exceed two‐thirds but do not exceed 90% of the minimum yield strength at temperature. The higher stress values should be used only where slightly higher deformation is not in itself objectionable. These higher stresses are not recommended for the design of flanges or other strain sensitive applications.

At this point, you may still be wondering why this is an issue. Well, the stress-at-a-point in a vessel component is not uniformly at or less than the allowable stress. In fact, the localized stresses can be substantially more - as much 3 times more (or even greater, in certain circumstances). So, in these types of situation, it is acknowledged that there will be higher stresses - and in the case of the "higher" allowable stresses, almost assuredly, the stresses will exceed yield.

(Of course, an appreciation of the full elastic-plastic stress-strain curve of these materials in needed, as they are usually quite ductile, yet exhibit a low proportional limit).

So, you need to understand the strain-sensitivity of your component. If some plastic deformation is acceptable/permissible, then you are very much welcome to use the "higher" allowable stress values for your design, and still be confident that your vessel will have an equivalent design margin against burst as any other vessel. And, based on experience, the "lower" allowable stress should be used if your component is strain-sensitive. In fact, you should probably do some calculations (FEA is NOT required) to demonstrate to yourself that the selected allowable stress is appropriate; sometimes for strain/deformation-sensitive equipment, even the "lower" allowable stresses are not low enough.

So, figure out, first, what the potential failure modes for your component are, and then select the allowable stress accordingly. And, of course, read the footnotes and the Appendices.

 

[RPRad]

Agreed. Aside from flanges assemblies etc, I have generally considered under what conditions the vessel is going to spend most of its operating life in determining which set of values to use.

 

[TGS4]

I have generally considered under what conditions the vessel is going to spend most of its operating life in determining which set of values to use.

Why?

 

[mechyp]

RpRAd

In section II part D

Recommendations given for not to use higherstess values for gasketed joint and other application where slight amounts of distortion can cause leakage.

Read general note G5.

Correct me if I am wrong

Thanks

Yp3135

 

[RPRad]

YP3135 - note G5 is the only guidance provided and it is clear on these points I was generally inquiring whether anybody had other situations where they would use the low stress

TGS4 - Other situations where I have used the low stress values over the high stress values are situations for example where the the design and operating temperature is close to max permitted by ASME (ie. SA240 316L operating around 800 F with design at 825 F...max permitted by ASME VIII is 850 F) with pressures in the 800 psi and up range). Perhaps this is an over design but I sleep better at nite having used the lower values!

 

[TGS4]

RPRad - while you are more than welcome to choose an allowable stress that is less than that required by the Code, it is neither required, nor economical to do so. However, it is your company's or client's money, and if wasting such permits restful slumber, I guess more power to you. I just hope that you are understanding all of the potential failure modes - when you have high temperatures, often you will see high temperature differentials, which result in thermal stress, which is often mitigated by thinner sections. I have come across situations where the philosophy of "when in doubt, make it stout" was actually detrimental, and lead to a premature failure.

So, think long and hard before you derate your material. You may be doing more harm than good...

 

[RPRad]

TGS4

That is a good point, I agree there are situations where stout is certainly not the best approach (cost has generally not been an issue at least in the instances I mentioned). Anyway what I take from this thread is that aside from leakage issues (and potentially stress sensitive areas) most use the higher stress values for the SS in most other situations...so that really answers my original question....in the reverse of what I expected