Hello Stressers!!!

As we all know the input for harmonic shaking force at elbows and change of direction for reciprocating compressor piping is the force amplitude (0 to peak) unbalance load.

My question is what is the fatigue stress output for this (0 to peak input of force)? Is it "zero to peak stress" or "peak to peak stress"?

The "peak to peak stress" allowable as per API 618 for CS material at a certain temp. limit (correct me if I'm wrong) is considered 26000 psi. Considering an ASME VIII Div. 2 FSRF of 4 and a factor of safety 2, the "peak to peak" stress allowable then becomes 26000/8 psi=3250 psi. This is one of the approach that I've seen in one of the acoustic consultant report in their "Mechanical Response Study" which can be considered conservative since an FRSF of 4 is the highest factor (weakest weld in ASME VIII Div. 2)plus adding a factor of safety of 2 for other uncertainties.

If in case the Caesar output for Fatigue stress is "zero to peak" then the allowable would still be divided into 2 (3250/2=1625 psi).

Kindly share your insight on this one. The easy way out is to use the 1625 psi as the allowable. But I'm thinking, if I'm imposing excessive conservatism in the piping system (which of course is good from safety standpoint) but may not be good from practical standpoint since pulsation forces may need to be reduced more from acoustic analysis view point (or if in luck maybe friction supports such as hold down will be enough to restrain pulsation forces in order to give way to thermal loads).

Any experience on this field is highly appreciated.

Cheers!!!

Harmonic analysis in CAESAR II will produce zero-to-peak results - amplitude.
You can see this by reviewing different phase angles for the same frequency analyzed. The program "recommends" a system snapshot at at the phase angle that produces the overall maximum deflection. Take another snapshot ("select frequency/phase pair") 180 degrees out of phase with the provided maximum. You will see, essentially, the same magnitudes with the opposite sign.

Thanks Dave for that input.1625 psi (zero to peak) fatigue allowable then is the right way to go.

Warm Regards,

Sorry Dave. Another way to approach this is to multiply the fatigue stress output from Caesar II by a factor of 2 in order to convert it to "peak to peak" stress in which I can directly compare to "peak to peak" fatigue stress allowable of 3250 psi?
Am I right in saying this?

Many thanks for your help and prompt reply.

Warm Regards,

Another method that can be used is to utilize ASME VIII Fatigue Curve for Polished specimen. At 1E11 cycles (approximately 158 Hz for 20 yrs) the allowed amplitude stress Sa=48.2 MPa (zero to peak). Since this is polished specimen need to divide by a factor of 2 which will give 24.1 MPa to take into account that ASME B31.3 uses girth butt-weld pipe as the basis for SIF's in Appendix D. Since this is a high cycle fatigue just make sure that D/t must be less than 100 or Appendix D is out of range already. Need some tools like FEA Tools to calculate SIF's for this case.

Cheers!!!