Flange check ASME VIII failed

Posted by: Andrii

Flange check ASME VIII failed - 04/25/16 01:27 AM

Hello everyone.

I have a model with socketweld flanges 1” (150# and 300#) and spiralwound gaskets. All flanges failed Peq check. Then I check them by method described in ASME VIII Div2 with next parameters:

- Flange material – A105N, studbolts material – A193 GR.B7
- Design pressure – 1650 kPa,
- Axial force – 9 N,
- Moment – 1 Nm,
- Gasket seating stress – 50 000 kPa.

Flange failed in seating conditions (radial flange and maximum average). After I change design pressure to 500 kPa, axial force to 0 N and moment to 0 Nm. Gasket seating stress is same – 50 000 kPa. Radial flange and maximum average in seating conditions not changed. So, I understand that specified parameters (in seating conditions) does not depend from forces and moments. How can I solve this problem? Could we manually change allowable stress in seating conditions up for 150%?
Posted by: KimTaeSung

Re: Flange check ASME VIII failed - 04/28/16 10:50 PM

Hi Andrii
If you change External Loads(Design pressure, Axial Force, Bending Moment), Longitudinal Hub stress, Radial Flange stress will be changed by calculating. But Bolting seating stress isn't changed.
If you change Gasket seating stress, Only Bolting seating stress will be changed by calculating.
In my opinion External Loads isn't related to Gasket seating stress and Bolting seating stress. And I think Gasket seating stress is not result value when you check the Flange failure. In my case before I check the Flange Failure, I could get the value from ASME VIII Div1 table for some types of gaskets.


I wish stress anlaysis experts solve this problem.
Posted by: Borzki

Re: Flange check ASME VIII failed - 06/06/17 07:01 AM

The reason for this is the ASME B16.5 Flanges were developed before 1937 where as ASME VIII calculation for flanges is developed in 1937, so sometimes some ASME B16.5 flanges does not pass in seating stress in bolt-up condition. This happens mostly in small diameter flanges like in the given example.

Any other opinion on this issue is greatly appreciated. This is one of the most common dilemma that stress engineers encounter.