Toward more consistent pipe stress analysis Presented are some guidelines in applying stress intensification factors to piping weight loading and small branch connections. It is hoped this information will alleviate the controversy and lead to standardization LC, Peng, Consulting Engineer, Houston Ixy Metis nxstaw msroRy, 1955 is a monumental year. Ins that year the stress range concept was formally recog- ied by the Code for Pressure Piping? as the basis for rating Uerinal expansion stress. Although the code ren expanded and clarified aver the years, there are stifi unsettled arguments regarding application of the code in certain areas. Tywo areas where inconsistencies still, exist are stress intensification factors for weight and thse stesely fondings and stress intensification factors for small branch connections. ‘These areas will explored along, with suggestions for applying the code. © Stress intensification factor for woight and other steady loadings. The stress intensification factors given in the code are intended for flexibility analyses. No specific intensification factor for weight and occasional loadings is mentioned in the Chemical Plant and Petro- eum Refinsry Piping Code.t Due to this tacit position of the code, piping designers are divided in actual prac tice, Some designers will apply the code stress intensficas tion Factors io all catogories of loads, while many other Gesigners tcncl to ignore the stress intensification factors muiplotely in steady lend analysis, One component acx ceptable to one designer can be rejected by others due to different opinions in the interpretation, Hyononannon Processive May 1979 © Small branch connections, The stress intensifica- tion factors given in the code for branch connections are derived from full size branch connections, These factors, although applicable to small branch connections, can become excrisively conservative for sinall connections on big pipes. Because of the apparent_overconscrvatism, designers offen ignore stress intensification factors at small branch connections. However, practices are never stent. For instance, it is ensy to sce that the stress tion due toa %-inch connection can be ignored in the analysis of 20.ich header, but for a 3einch cone nection, the factors to apply will differ among designers. Stress intensification factors given in the code for branch connections can be too conservative for small connections on large pipes 207TOWARD MORE CONSISTENT PIPE STRESS ANALYSIS Fig, St hing a bar with 3 smati naio ‘These two examples are related to the application of stress intensification factors. Apply or not to apply very often means several times difference in the allowable loads, ‘These are determined solely by designers’ personal pri ferences and inclinations. A more consistent approach needs to be developed and adopted. STRESS iNTENSIFICATION When a structural member i siretched, stress in the main uniform section can norznally be caleulated by sim- ple formulas, but the stress in a locally notched or sti fened discontinuous section is either very complicated or impossible to calculate. For practical design purpose, stress at the discontinuous section is estimated by applying 2 stress intensification factor over the siress calculated at the main uniform section, ‘This stress intensification can be derived theoretically or determined by test. At a structural discontinuity, stress intensification can, be quite different for different types of loading, Big. 1 shows a long rectangular bar with a small hole iu che middte of the section. At Section A-A outside the influenice of end fixtures and the hole, the stress is-uniforinly tributed at a magnitude of S=F/(bt}. But at Section B-B, due to discontinuity in strain flow, the stress is un- evenly distributed. A maximum stress, Sua, of about three times the uniform sterss occurs.at the edges of the hole. ‘The stress decreases very rapidly at points away from the edge of the hole. Theoretically, the hole has created a stress intensification fnctor of three, but its signifiéance is different for diferent materials. Por a brittle material such as glass, the hole will degrade the bar to one-thied its original strength because it fails 5 soon as the maximum stress reaches fellure stress. Piping materials, on the other hand, an: normally very ductile, rand 2 considerable ameunt of yielding takes ince before the member fails. With ductile materials the stress ine rensiication needs to be interpreted in two different categories, namely steady and cyclic. Steady loading. Under steady loading the highly tocal- ied stress concentration: will be redigtributed to the at Jaoent urea once the local stress reaches the yield point, Eventwally the load will spread evenly to the whole cross- section hefore the Dar falls "Phe innportant stress is the sibow 1 ier bene Equtvaent 4) bows Fig. 2—Equivelent elbows rodistributed stwess prior to the failure, Since the se distributed stress is casentially the averaye stress, the stress intensification factor for steady loading is es ee my Fi] (bya To = ai } which is entirely due to reduction of tke erow-sectional area, Cyclic loading, Under cyclic lowding the member fails dus to fatigue. Since the primary measure of fatigue faite tre is the local strain range per cycle, redistribution of stress due to plastic How is not very iniportain, Therefore the stress intensiReation factor Jor cyclic lending is ie Soup 2 which is the measure of the mmtsineuns focal stain. Siar is the maximum equivalent elastic sires futher than the actual stress. Elbow stress intensification factor. In piping stress analysis, the elbow sins intensification factor i particu larly Important not only because the elbow ednstitutes a major portion of the system but als ixcause it i the basis for deriving the stress intensification fietor for other component shapes. For instance, Markl succesfully used elbow analogy to correlate his fatigue test results on tees and initer fends, Using the eqnivalent elbows as shown in Fig. 2 and making adjestments for actuat crotch radi and thickness, a set of stress intensifieation factors was constructed using a single Rexibility characteristic para. meter, . A detailed discussion on elbow characteristics is denoficial in undessiandiag the werd of all components. « . ‘An clbow behaves very dillerently from a stiaiyht pipe in resisting bonding moments, When a straight pipe is bent, its erose-section remains cheular and stress itereases finearly-with distance [rom the neutrul axis, Hoavever, when ap elbow is beet as showa in hig. 3, the cross section deforms to an oval shape, This evaiization is doe to Joss rigidity at extreme Abers in the tangential to direction, and less energy being weeded for the elbow 10 assume an oval shupe tian to maintain a cirear cross section, Top and bottom portions of the pipe wall simply duekte in to escape froin carrying diet” proper share. sectlan ren. Congitaial ‘tresees Fig, 3Stress doformation of an elbow of the lad. The bending moment is resinted essentially by the shaded effective section, ‘The, magnus stress point is shifted from Point A to the effective exireme Point B. As the eross-scetion ovals, a loca bending stress is also produced around the circumference, ‘The maximura circumferential stress occurs at Point @ where the radius of curvature Is the smallest, Mathematically the maximum longitudinal stress and ciewmsferential stress can be calculated by using the following stress intensification factors: = owas . aon } inplane bending (3 Be = 1.001% We iSOphYS ‘The experimentally measured distributions of the longitadinal and circumferential stresses of a_$0-inch pipe elhow subject to in-plane bending’ are shown in Pig. Masimisn circumferential stress is normally greater than the maximum longitudinal suess, However, the nature of the tn stresses is quite different. The lengi- tudinal stress is a membrane stross working directly against he manent, wily the circumferential stress isn. skin bonding stress resulting from local deformation. outsplane beading {4} Code stress intensification factors. The stress in- tnsifcation factors given ia the cadet are intended for izeraal expansion and other displacement loads. The ature of thermat expansion foad is different from that of weight and other sustained Toads. Thermat expansion is self-limiting, {¢ is a strain controBed bading such that cover the strain reaches a paint large enough to com pomsate for the expansion, growth stops regardless of the actwl stress ceveloped in the system, Tt can not mally cause any structural damage in one single appl (ion, but aa cause fatigue failure through repeated ox. pansion and contraction cycles. ‘Therefore, for evaluating, final expansion, the strass intensifeation factor is de- lermined hy thr ritio of the steess causing faihure over a given amber of eyeles in a straight pipe to the stress causing fale at a component subject to an equal nutr- bor of stress aycles, Code stress intensification factors are suress intensification factors in which the loval pene strass is governing. Hyorocannon Processing May 1979 Theoretically these intensifications ure equal to the maximum stress intensification existing in any region and direction within a component, In an elbow, for instance, the circumferential stress intensification factors 1.80/4% and 150/k¥ for in-plane and out-of-plane bendings, ively, should be used. However, intensive fatigoe ious componcnts* have shown that by using unity as the fatigue life of girth welded or clamped pipe, the effective stress intensification factors of elbows in bending fatigue were about half the theoretical value, By dividing the theovetical factor by two, the code stress intensification factor for elbows is as fellows In-plane stres intensification factor i = 0.90/18" (3) Octofeplane stress intensifiention Sactor 75) 6) “The stress intensification factors for other components are derived by using elbow analogy correlated with test esalts, STRESS INTENSIFICATION FACTORS FOR WEIGHT AND OCCASIONAL LOADS No stress intensification facior is explicitly stated in the Chemical Plant and’ Petroleam Refinery Piping Code for weight and occasional fonds. Weight and wind are ustained Ioadings, They are not self-limiting, and always reqqire a static equilibrium between the stress developed in the componcat and the load applied. Once yield point ox collapse load is reached, the component will fail rogardleis of the amount of deformation that has occurred. ‘Therefore, the stress to be considered in weight and other sustained Toadings should possess the following charac teristios: W The stress is in a dincetion directly agninst the loading. Oniy the stresses acting npuinst the load ate Inadlacarrying stresses D> ‘The stveas is the average stress weress Ue wall thick- ness. ‘The average stiess is actually the remaining stress ‘available for external equilibrium alter the fteyaal mu tual cancellation. Theory and experiment indicate the same code stress intensification factors intended for flexibility analysis should be used in weight, occasional and other sustained load analyses 209