Half-pipe jacket weld

[roscoe3588]

I am currently reviewing calculations for a new batch reactor designed in accordance with ASME Section 8 Div 1. The reactor is subject to thermal cyclic loading. The client suggested the possible use of "full" penetration welds of the half-pipe to shell to improve fatigue. The FEA and calculations submitted indicate a service life of about half of that when fillet welds are considered due to the fact that the larger area of a full penetration weld versus that of a fillet weld causes higher stresses. Thus when using a master S-N curve, less cycles are predicted. If we follow the ASME code and use penetration welds for the thermal cyclic loading, the predicted service life will be significantly less. Am I missing something in the interpretation or application of the code? Any suggestions?

 

[MJCronin]

"the larger area of a full penetration weld versus that of a fillet weld causes higher stresses."

Something is not right here ....

Both ASME B31.1 and B31.3 caution against using fillet welds in thermal cyclic service. Full penetration welds are preferred.

MJCronin
Sr. Process Engineer