When testing a pressurized pipe with off-standard flanges, often times the materials available to fabricate a blind flange of the correct size are limited to what is around.
I've made it a habit of referring to Roark's formula's for stresses and strains, and going to the table for large deflection circular plates for these. I use the area inside the bolt circle. They have three boundary conditions available:
- Simply supported (neither fixed nor held)
- Fixed but not held (no edge tension)
- Fixed and held.
Bolted connections are unusual - especially since I have a full face flanges. I've designed to both simply supported and fixed but not held. The simply supported works, the fixed but not held does not. However, due to the flatness out to the OD of the face, it seems like the rotation of the edge should be 0, so it should be fixed but not held. I'd like to make these thinner, so it would be nice if someone can demonstrate an accurate boundary condition (or perhaps that 80% of the fixed + 20% of the simply supported would be conservative via FEA). I'd like to see better research into the mechanics of design for blind flanges (which is rarely covered in the context of the Boiler and Pressure Vessel Code, or elsewhere).
The main question is: what boundary condition for the edge should I use?