ASME BPVC VIII-2 2010 fatigue screening

[TerryR1]

I have a pv made from 17-4ph ss subject to 2000 cycles design life.

Equation 5.18 (p5-16) says I can skip the fatigue analysis if my number if cycles is less than "N(C1*S)".

S is the allowable stress (58ksi) from Sec II. C1 is 4 from a table. So I am checking to see the number of allowed cycles at 4x the allowable stress. Turns out I am limited to ~15 cycles, regardless of the actual stress or pressure I put in my vessel.

What am I missing? Thanks!

 

[fegenbush]

It seems that you have it right. Current design to Section VIII Division 2 is setup for FEA of almost all designs, as you have seen. From your results, you have three options: change the design basis to one in which fatigue analysis is not required, perform the fatigue analysis, or switch the design to Division 1.

 

[TerryR1]

Thanks fegenbush.

 

[TGS4]

I would second some of fegenbush's response, with one very important caveat: do not think that gong to Div 1 is a panacea. UG-22(e) tells you that you must consider cyclic loads - it just doesn't tell you how. You would be making a very serious mistake going over to Div 1 without doing the fatigue analysis.

Also, Div 2 is NOT set up for FEA for ALL designs - the design-by-rules of Part 4 is set up to not have to do advanced analysis. If you have fatigue - and most failures that I have investigated fail do to cracking from cyclic loading - then there is no "end-run" around the rules. Do the fatigue calcs. You use a high strength material like 17-4PH to take advantage of the high allowable stress. However, 17-4PH is not similarly invincible to fatigue - and by having the vessel in a highly stressed state, and then cycling that, you are practically begging for a fatigue failure. DO THE FATIGUE CALC!

 

[TerryR1]

Thanks TSG4, I agree and will. What was surprising to me was how much effort it took to get that 15 cycle limit, and that the screening is independant of stress or pressure in the cylinder.