This level of overstress is good cause for concern. A full fatigue evaluation would be good but as pointed out getting sound meaningfull data is difficult at times.
B31.3 & .1 uses a base load of cycles of 7000. This is a very high figure for most systems and we have speculated that many a marginal design / installation has been saved because the N the system actually sees is lower.
However low cycles at high stresses gets tough because the scatter of the data can get you in trouble so many people try not to go N=100.
If you have the $$$$ API 579 is helpful.... you may want to drag some API 579 experts into the fray.
The problem with modifications at this point is the modifications will reset the clock on the thermal strain energy in the system and if things had settled down they will have to undergo a new shakedown.
Or if you get the old "Its worked OK for 10 years it must be OK" you make your recomendations and walk away....
Best Regards,
John C. Luf
Cleveland Ohio U.S.A when I'm not in Austria
Member B31.3, Piping Engineer - Stress Analysis, Pipe Supports, Component Design, and Surge Analysis... and according to my daughters master of unimportant trivia
[Non-text portions of this message have been removed] Received on Fri Feb 24 20:22:00 2006
This archive was generated by hypermail 2.1.8 : Mon Oct 27 2008 - 20:24:10 EDT