Hello Fellow Stressers!!

Expansion stress range in the fatigue analysis will use the Fatigue Curve in ASME (which is given in Stress amplitude) to be multiplied by 2. Essentially, the alternating stress being 1/2 of expansion stress range.

But how can we take into account the mean stress & loading ratio because normally, the loading ratio will not be zero at some point in the cycle life. Say the tie-in temperature is 70 deg. F and design temperature of 170 deg. F and minimum ambient temperature of 27 deg. F. So approximately the loading ratio (ratio of 2nd load to 1st load) will be (27-70/170-70) equals -0.43 and there will be some shutdowns at varying ambient temperature, which will give different loading ratio also at some point in the cycle life.

Would it be prudent to use the ASME Fatigue Curve (Stress Amplitude) to compare directly to stress range in lieu of a more detailed fatigue evaluation (considering mean stress effects & loading ratio) or this is already accounted in the fatigue curve itself.

Your expert advise is highly appreciated.

Warm Regards,

Hi Borzki,

Quote from ASME BPVC VIII-2 Sect. 5.5.1.4:

<< Stresses and strains produced by any load or thermal condition that does not vary during the cycle need not be considered in a fatigue analysis if the fatigue curves utilized in the evaluation are adjusted for mean stresses and
strains. The design fatigue curves referenced in 5.5.3 and 5.5.4 are based on smooth bar test specimens and are adjusted for the maximum possible effect of mean stress and strain; therefore, an adjustment for mean stress effects is not required....>>.

So, the common employment of smooth bar fatigue curves given by ASME VIII-2 part 5 do not require mean stress on cycle consideration.

Best regards,

Thanks Dorin for that precise information regarding mean stress. It clears my mind now. How about loading ratio?

Warm Regards,

It's ok now Dorin. As I understand loading ratio is also related to mean stress. And since the fatigue curve itself have considered the effect of mean stress different loading ratio can be catered but with consideration to the cumulative usage for each cyclic stress (in case the system has non-constant amplitude).

Thanks,