ASME B16.5 Hydrotest Pressure Limits & Yield Strength 1

[gwalkerb]

Note that everything below is based on piping systems built to B31.3 with B16.5 flanges.

Our QC and hydro-test group had occasionally in the past accidentally taken hydro-tests to pressures higher than specified. Although there weren't any visual indication of failure (i.e. leakage or deformation), they passed the details of the test back to us in engineering to validate if the materials were still usable, or if they needed to be scrapped.

The flanges are typically the limiting component in our piping systems, and our standard hydro-test pressures are based on 1.5x the maximum design pressure of B16.5 flanges at 100F. The methodology that was explained to me to validate over-pressure during a hydro-test (by someone who no longer works here), was to use the formulas from B16.5 Nonmandatory Appendix A section A-2, but calculate a pressure using 100% of the yield strength at 100F instead of 60%. If our hydro-test pressure didn't exceed that value, then we could be confident that we didn't exceed the yield strength of the material, and didn't create any plastic deformation.

To provide further protection, and avoid ever damaging flanges, we've decided to use relief valves in our testing system, set appropriately for each flange class and material group. So I've been tasked with sizing and specifying these relief valves.

Using the methodology above to determine relief valve set-pressures works fine for group 1 materials (e.g. A105N, A350-LF2 CL1), which is the majority of what we fabricate. We do sometimes use group 2 materials (A182-F316, A182-F304, A182-F316L, A182-F304L) but when I was doing the same exercise for group 2 materials, I realized that hydrotesting a group 2 flange at 1.5x the design pressure would then exceed the calculated yield limit. This arises as group 2 and 3 materials follow the same formula for determining rating, but use a yield strength factor of 70% instead of 60%. So a hydrotest at 1.5x design pressure which is based on 60% of yield strength would then occur at 90% of yield strength. But a hydrotest of a group 2 or 3 material with design pressure based on 70% of yield strength would then occur at a pressure equivalent to 105% of yield strength, which implies that the flanges could potentially be plastically deforming.

I'm not sure how to reconcile this information, and set relief valves appropriately for group 2 or 3 materials. Therefore I have a couple questions:

1. Is the methodology described above of using a 100% factor on yield strength, but still following the A-2.2(a)(1-3) formulas a valid way to determine a limit on hydrotest pressures?
2. If this methodology is flawed, what is the proper way to determine the limit of hydrotest pressures?
3. If this method is sound, how does that work with group 2 materials exceeding their yield strength based on the 70% x1.5?

 

[moltenmetal]

Excellent question- perfectly asked too. You've given all the information necessary to get the group to help you solve it. I don't have an answer for you but know there will be people here who can provide the needed advice- and I'll be reading those answers with interest.

 

[XL83NL]

I concur with MM. I dont have the answer for you, but just to be sure: I assume you are aware of para 2.6 and it's contents? It seems that that specific section provides the validation for your method, however it just doesn't match with the technical background provided in Appendix A para A-2.3 which you use.

 

[TGS4]

Agreed on the above re excellent question. I, however, feel somewhat qualified to answer...

There are a couple of things that you need to first understand regarding what is casually thrown around as "yield". Specifically, this is referred to as the engineering yield stress. This is determined from a full stress-strain curve using the 0.2% offset method. That means that even at the engineering yield, there is already some permanent deformation. A long time ago, our colleagues decided on 0.2% for repeatability. However, many stainless steels have a very gentle stress-strain curve, and so it has been decided that it would be acceptable to a slightly higher value of permanent strain.

In general, the criteria for determining whether a hydrotest pressure was excessive is gross deformation and/or leakage. If neither occur, then your flange is fine. My justification for this statement is that the calculations are based on minimum-specified properties and your flange is made with actual material that is likely stronger. However, I understand your concern for calculating a maximum test pressure and setting relief valves accordingly.

To answer your specific questions:
1) A-2.2(a)(1-3) is one way of performing that calculation. There are others.
2) Not flawed, but not necessarily optimal.
3) Exceeding the engineering yield during a test for austenitic stainless steels is perfectly acceptable. Even in the piping Code, the allowable stress may go up to 90% of this engineering yield, and so a x1.5 hydrotest will go to 1.35*Sy. This is all considered acceptable. You may need to do some additional calculations to determine the actual permanent deformation, but you won't find those calculations in B16.5. Personally, I would use EN-1591, but those calculations can be daunting for a first-timer. Just don't use the VIII-1 Appendix 2 calcs. They just ar not appropriate for this type of work.

 

[gwalkerb]

Thanks TGS4. I will look into EN-1591. I'm not familiar with that particular document, but from some googling it appears to be a British standard? Is there anything within the ASME or API realm that gives alternate methodologies? (as you mentioned I'll stay away from Section VIII, although I'll probably take a look for context).

While I don't have an issue referencing codes outside of the ASME realm, it would be preferred to follow something that our customers and jurisdictions are more likely to be familiar with. But if EN-1591 gives good results, I'll be prepared to back up my usage of it.