Hi Dave,
Perhaps I have mis-interpretted your previous response. And I agree the second approach of doubling the SIF's is a "sloppy" approach but I thought that was a way you were suggesting. Can you advise how you would perform a fatigue analysis using Caesar II adopting the approach of DNV-RP-D101 which proposes PD5500. In particular how would the curves of PD5500 need to be input to Caesar II as the PD5500 curves are based on Stress Range vs Cycles? How do we ensure that the "actual" Stress values (taking into account the correct SIF's not the Code SIF's) are compared with the correct fatigue curve?