Hello,

I used the CAESAR flange check module, I noticed that the allowables for the seating condition & that for the operating conditions are different.
I used 1.5 for the allowable multiplier. The allowables for the operating condition was increased by 1.5 but not the allowables for the seating condition.
Please share some light on this.


Thanks,

Note, the numbers below are for example only and don't necessarily represent an actual situation.

When the flange is first assembled, and before the pressure is applied, the temperature is at ambient. This is the gasket seating condition where the bolts are tightened, subjecting the flange to a bending moment from the bolt load. Once the piping system with the flange components are pressurized with hot fluid the forces and moments on the flange change. Not only is pressure introduced, but the temperature likely rises as well. This is the operating condition. The flange has to be analyzed for the gasket seating condition when the flange pair is first assembled, and additionally for the operating condition.

The design stress of the flange material may be different at ambient temperature from that at operating temperature, so both design stresses must be considered. The same is true of the bolts.

Now consider the ‘Flange Allowable Stress Multiplier’ and the ‘Bolt Allowable Stress Multiplier' (ASME Section VIII DIV 2). There are three stresses that are computed: Sr (radial stress), St (tangential or hoop stress) and Sh (hub stress). Using stress multipliers increases the allowable stress by the ‘Stress Multiplier’ factor for St and Sr. This is a violation of ASME Section VIII, Division 1, Appendix 2, but not necessarily a violation of B31.1.

Consider the reason why high flange stresses would be allowed. Normally when flanges are designed to ASME Section VIII, Division 1, the maximum allowable bolt stress at initial tightening (gasket seating) is around 25,000 psi. We would have to limit the bolt stress to this value and at the same time not exceed the maximum allowable stress for the flange. However, when flanges are assembled in the field, the artisan does not see the calculations for the flange design, and quite frankly he does not care. He usually has a procedure that tells him he must apply a given bolt torque to achieve a bolt stress of 50,000 psi (not the ASME allowable value of 25,000 psi). Somehow the flange survives this brutal treatment. In the case of the ASME code, the design function and what happens in the field is ignored. B31.1 seams to recognize this anomaly.

If you look at the help screen for the ‘Flange Allowable Stress Multiplier’ field, you will see this note:

“This increase is implied in B31.1 Appendix II Section 4.2.3 when it states that the longitudinal hub, tangential and radial stress allowables are equal to the yield stress at design temperature, which is essentially 1.5(S). Users wishing to apply this increased allowable to the radial and tangential operating ASME flange allowables should enter 1.5 in this field. Note, prior to the 1992 edition of the ASME NC code, NC paragraph 3658.1(d) also stated that the tangential and radial stress allowable could be increased 50%. The 1992 edition of NC eliminated this increase on these allowables”


The note above indicates "yield stress at design temperature," meaning that the seating condition is not subject to this B31.1 paragraph.

Now, using a multiplier of 1.5 x S brings the stress up to the yield stress which is the theoretical limiting stress, because plastic deformation occurs when yield stresses are surpassed. However, generally much higher stresses are applied in the field, strongly suggesting that the design of flanges is actually very conservative (I am not recommending increasing this stress multiplier though).

Hello Loren,

Thank you for giving me a free lecture.
As a follow-up question, look at the attached file and see that the allowable seating stress for the longitudinal hub stress is increased same as the allowable operating stress. While the other seating stress allowables were not increased. Why is this so?

Also, Is it logical to think that the flange assembly should be stronger or at least equally strong at seating condition than at design condition ?

Thanks & hoping for your guidance,

Julius,

The hub stress is a bending stress where the other two are not. In someone's interpretation of B31.1 (I am not sure yet who made this decision) the multiplier can apply to this. As for the strength of the flange assembly I would tend to agree with you, but B31.1 specifically states at design temperature and not in the seating condition, which is why I am surprised we are applying the multiplier to the hub stress in the seating condition. Remember that ASME Section VIII does not allow this multiplier.

The allowable stress multiplier of 1.5 for bending stresses comes from the physics of plastic hinges. It is not a "decision" by the B31 committee, it is reality.

You can go back and read up in Timoshenko or some other basic reference for simple beam theory to see what happens to a round section subjected to a bending load sufficient to produce plastic strains if you want to understand this. Better yet, dig out your simple beam theory text and work through the problem yourself.

Craig,

The allowable stress multiplier of 1.5 for bending stress, if someone uses limit analysis theory for circular c/s will not come come to 1.5, it will come to around 1.33.So I believe that is has to do ( besides limit analysis), a factor that the stress does not exceed y.p.

In fact if inverse of 2/3 i.e. F.O.S on allowable stress , is taken,it comes to 1.5.

Best Regards

Thanks everybody for sharing your thoughts.
I know through this forum that there are modern methods of flange analysis, but I have no knowledge yet on them, that is why I have to use this traditional, proven, conservative approach.

Regards,