B31.1 para 104.3.1 weldolets 1

[sjrfc2]

104.3.1 (C) indicates that calcs don't have to be performed on reinforced branch connection as long as the fitting and the weld meet 104.3.1 (D). This seems to steer the user into doing the reinforcement calcs. It seems a little contradictory but my question really has to do with what is the required weld size.

With the MSS SP-97 fittings I think there is the assumption that you weld to the outside of the bevel but you don't see that called out. This would be the assumption with matched pipe and fittings. I'm really interested in overmatched fittings.

We have standard weldolets and use them on thin-walled piping (Sch 10). If we weld out to the end of the bevel, we run the risk of collapsing the pipe due to weld shrinkage. Pipe Fabrication Institute ES-49 discusses this.

We'll have plenty of reinforcement because its such a beefy fitting but taking that to the extreme, you'd need very little weld (really no weld per the calc) to meet Code. This obviously isn't the intent but we're trying to establish what weld thickness is required to transfer stress to the fitting (where we're taking credit for the reinforcement) as well as provide a good weld for mechanical loading since the reinforcement calcs are only for pressure. One thought would be to limit area of reinforcement parallelagram to OD of weld but I figured I'd see if anyone else had run through this.

Thanks in advance for any thoughts on this.

 

[Duwe6]

See B31.3-2006 Fig. 328.5.4D (1) Does not address O'lets / Integrally Reinforced Branch Fittings.

B31.1 Fig 127.4.8(E) does address these fittings, sort of. These dwgs are based on Half-Couplings.

My interpretation of the intent of ASME is:
That the O'let weld be at least as thick through the fillet [thus 0.7 x leg size] as the wall of the pipe it is being attached to.
A full-penn weld is required, and the inside of random fittings should be visually verified that there is indeed a root present.
Finally, that the weld 'fair in' to the pipe smoothly, with no 'rolled' profile to make stress-risers. Welding more than the above size is acceptable, but 'flushing out' an O'let weld is not required.

 

[DSB123]

Weldolets have specified Stress Intensification Factors applicable to them as per the Code. In order to use these SIF equations the welding of the weldolet to the pipe has to meet the manufacturers requirements (i.e. be fully welded out to the bevel) otherwise you cannot use the Code SIF's. The problem then comes as to what SIF values are applicable = You cannot use the Code values.