Hi Ed and all,
By way of clarification (less someone get the wrong idea), I would like to revisit the B31.1 treatment of SIF's when applied to Sustained Stress calculations. The bending moment is intensified by multiplying it by the SIF and this product is divided by the section modulus to arrive at the intensified bending stress (SLB )
(SLB ) = 0.75 * I *M / Z
and the product of 0.75 * I, MAY NOT BE LESS THAN 1.0 (B31.1 Para. 104.8.1). Note that Ed is referring to the SIF MULTIPLYER option in
CAESAR II as being 1.0, 0.75, or 0.0001. The resulting product SIF per se, still can not be less than 1.0.
We tend to assume that everybody knows that the SIF will never be less than 1.0 but every now and then, in seminars, we find out that it is really not common knowledge. But the issue for discussion is whether to use SIF's at all in calculating Sustained Stress.
It is because B31.3 (Process Piping) does not provide an explicit equation for calculating the Sustained Stress due to pressure and weight that the piping design community turns to the B31.1 Power Piping Code for guidance. The B31.1 Code prescribes the inclusion of the SIF's in the equations for Sustained and Occasional stress.
Recently an inquiry asked B31.1 if it was the intention of an older edition of the B31.1 Code (USAS B31.1, 1967) to include SIF's in the calculation of Sustained Stresses. The inquiry resulted in the development of interpretation 34 - 2:
Question: Does USAS B31.1 – 1967, paras. 102.3.2(D) and 102.3.3(A), require consideration of intensified longitudinal stresses……… produced by weight and other sustained and/or occasional loads? (truncation is mine).
Answer: No. USAS B31.1 – 1967 does not require that the stress intensification factors (SIF's) listed in Appendix D, or fraction thereof, be applied to the longitudinal stresses produced (sic) by weight and other sustained/occasional loads. However, as stated in the Forward, the Code never intentionally puts a ceiling limit on conservatism, and a designer is free to specify the more rigid requirements as he feels they may be justified.
If you look at B31.3 interpretation 1–34 (2/23/81) you will get the impression that B31.3 did not want you to use SIF's for calculating sustained stresses. If you look at B31.3 interpretation 6–03 (12/14/87) you will get the impression that B31.3 allowed you to use SIF's for calculating sustained stresses at your option.
Take a look at John Luf's article in the
January 2001 issue of the COADE newsletter. John (one of our esteemed colleagues in stress analysis) voices his opinion that in regard to Sustained loadings and resulting calculated stresses, “the SIF is not applicable to sustained loads”. John's argument is informed and compelling.
Ed's observation that the SIF's were developed to “adjust” the calculated (by beam theory) expansion (secondary) stress to include the effect of bending cycles is astute. So, how does the application of SIF's to Sustained Stress calculation have any claim to legitimacy? After all, the SIF simply indicates the ratio of the number of cycles to failure of a straight piece of pipe (with a girth butt weld) to the number of cycles to failure of a specific piping component (e.g., elbow, Tee, etc.) And, the word “sustained” seems to preclude the consideration of cycles. Well, the Markl fatigue tests certainly did show that stresses are “intensified” in the irregular geometries of the tested components. Some will argue that the pipe does not know if the stress it is feeling is due to expansion or weight – it just knows that it is stress. It is also argued in some discussions that since B31 Codes do not address explicitly pressure cycles, it is appropriate to use SIF's to “allow” for additional conservatism in calculating the Sustained Stress.
There are some other issues that might help a stress analyst to choose what option to use in regard to SIF's. Remember that the SIF's were developed from tests on exclusively NPS 4, standard wall, ASTM A-106, Grade B carbon steel pipe and components at “room temperature”. The results of the tests were “extrapolated” down to NPS ½ and up to NPS 72 (?????). It is to be judged by the engineer to what degree these data are useful when applied to other materials, other diameters, et. al. Carbon steels and stainless steels have significantly different material properties in fatigue. It is often argued that at higher temperatures, the lower allowable stresses developed for stainless steels within the creep regime are conservative enough.
So, what are we to do? Well, it is certainly clear that the B31.3 Code does not MANDATE that SIF's be used for calculating Sustained Stresses (i.e., they did not “write it into the Code”). The decision to use the B31.1 equation IS a “common industry practice” but it is up to the engineer to determine how to use it. It ain't for nuttin' that
CAESAR II gives you some latitude via multiplier options. If you chose to always use 0.75 times the (Appendix D) SIF's in calculating Sustained Stress, you won't be alone. And on the other hand, if you chose to use no SIF's in calculating Sustained Stress, you won't be alone. Read John Luf's article again.
The above is not necessarily the opinion of ASME International nor is it the opinion of any Code Committee. It is offered as only my opinion in an effort to stimulate discussion among practicing engineers.
Best regards, John.