The short answer is yes. But it is important that you understand why this discrepancy is occuring.Most standards/codes are written utilising/assuming nominal stresses. Any accountance for peak stress in a structure is handled by predetermined Stress Intensity Factors, the sort you will find tabulated in Roark's (or a similar engineering textbook). These factors do not always account for all potential geometry configurations, as that pdf highlights regarding d/D for branch connections, and sometimes are simply not accurate for the stock standard geometry.These Stress Intensity Factors have been in use for decades (probably more than a century?) and most standards are written using them as FE techniques are much younger. Because of this, it is always a challenge to correlate an FE model with standards. The first check is usually to take the nominal/normal stress from your model, see if that complies, then use any of the SIFs that the standard references on the nominal stress. Then if there are discrepancies between the peak stress from the code and your model, try to understand why.

Hi Ben, thanks in advance for your explanation.nIt is gave me a clear vision about this issue.nBut the other problem still there when we discuss and generate the report to the client.nI think because not every client is understanding the history of the standards and the SIF.nDo you have any experience about this?n