Dear every stress analysis colleague:
As we know, the SIF value in AppendixD of ASME31.3 is limited by the ratio of d/D and the ratio of D/T.
So during piping flexibility analysis, to the critical pipe I want to give a more exact SIF value to the tee. I use Nozzlepro to get them. the value form Nozzle from always is less than that from CII.Even I get value less than 1.
First one:I feel confused why there has vlaue less than 1 for SIF?
Second one : in note 12 of Appendix D, if 0.5<d/D<1, The value is non-conservative, it mean the actual SIF is more than that from B31.3, But
from Nozzlepro , I still get less value?
third one: the SIF to branch is always bigger than the SIF to header,
how to select which should be used?

Thanks!

stress river,
The ASME B31.3 Code has many items established by "rules". The SIF was established from lab fatigue tests and then generalized to other sizes and types of fittings. CaesarII uses the Code "rules". The ASME and B31.3 Code tends to be more conservative. It is your decision to use your results from NozzlePro or to use the Code SIF in CaesarII. Which one do you believe is correct and accurate? Which one do you want to risk using?
An engineer or designer is not restricted from doing a more accurate or precise analysis or design using other methods - he just has to document and prove out the design by testing, etc. If the sizes and thicknesses are beyond the Code design tables, or it is beyond sizes /wall thickness of past designs, then it could be time for some prototype tests.

Well said Richard.

Also consider that the ASME B31 Codes for Pressure Piping include SIMPLIFIED rules.

Caesar II uses the SIF equations that are found in the Codes.

The original (Markl) testing that was done to establish the S-N curves that were "matched" for the SIF equations was all done on NPS 4, schedule 40, carbon steel pipe (no other sizes, schedules or material). The results were extrapolated. Consider the range of pipe to which we apply these simple equations - down to the smallest and up to very large sizes (note the warning about the applicability of SIF's to pipe components with large D/t ratios). The B31 Codes tell you that YOU MAY USE THESE SIF's IF YOU DO NOT HAVE BETTER DATA. If you go to more rigorous analytical methodologies then you DO have better data for the size and material of the component. So, if you have better data, override the default (B31) SIF and document it.

It is kinda analogous to comparing a sun dial to an atomic clock.

Regards, John

'It is kinda analogous to comparing a sun dial to an atomic clock.'

Breen-Sir,

There is no doubt that you provided an apt comparison; but shouldn't there be some caution, too! I have seen some openly declared 'non-calibrated quartz' digital clocks giving funny accuracy compared to hand-wound ones.

Whevever we are having less conservative soutions based on FEA analysis, we need to safeguard ourselves with formal documentation complete with review/checking & approval. In the process, we can get accurate analysis with analyst's risk hedged, too! it's a fact that FEA results can be willingly/ unintensionally manipulated.

regards,

sam

Why I am against mass scale use of FEA without proper human resource is that I have witnessed dangerous deskilling of such jobs.

In innovative terms of today, the nature of these jobs are mapped with the minimum skill required to meet the demand of reduction of cost & employee attrition. It is not unusual to find a class X tea boy promoted to be a FEA analyst as FEA analysts are costly & switch jobs somewhere. How can we trust his print-out!

regards,

sam