Every part of API 579 has rules addressing the case where a defect is too close to a “Major Structural Discontinuity” (MSD). But I’ve noticed in the API 579 training I’ve conducted for the last 15+ years, that most of the people in the course are unclear on exactly what constitutes an MSD. In this column, I’ll discuss the MSD: what an MSD is, and why it is so important in fitness-for-service (FFS) assessments.
Let’s Start with a Few Definitions
API 579-1 (2016), paragraph 1A.55 says:
“Major Structural Discontinuity – A source of stress or strain intensification that affects a relatively large portion of a structure and has a significant effect on the overall stress or strain pattern of the structure as a whole.”
The ASME Boiler & Pressure Vessel Code (BPVC), Section VIII, Division 2 uses the exact same definition, except that it calls them “Gross” structural discontinuities, instead of “Major” structural discontinuities. We know that “Gross” and “Major” structural discontinuities are the same because API 579-1 provides another definition that clarifies that point:
API 579-1 (2016), paragraph 1A.39 says:
“Gross Structural Discontinuity – Another name for a Major Structural Discontinuity.”
A slightly different definition is provided by the nuclear code, BPVC, Section III, Division 1, “Rules for Construction of Nuclear Facilities.”
Section III, Division 1, paragraph NB-3213.2 says:
“Gross structural discontinuity is a geometric or material discontinuity which affects the stress or strain distribution through the entire wall thickness of the pressure retaining member. Gross discontinuity type stresses are those portions of the actual stress distribution that produce net bending and membrane force resultants when integrated through the wall thickness.”
These definitions of “Major” or “Gross” structural discontinuities can be compared to the definition of “Local” structural discontinuities (LSDs) defined in Section VIII, Division 2 and API 579.
API 579-1 (2016) & BPVC Section VIII, Division 1 have identical wording and say:
“Local Structural Discontinuity – A source of stress or strain intensification that affects a relatively small volume of material and does not have a significant effect on the overall stress or strain pattern, or the structure as a whole.”
Comparing the definitions of “Major” or “Gross” to “Local” structural discontinuities, the main difference is that MSDs are defined as affecting a “relatively large” portion of the vessel, whereas LSDs affect a “relatively small” portion of the vessel.
Personally, I find this definition a bit too imprecise for engineering purposes. I prefer to think of this in terms of the definition from Section III which says the MSD “…affects the stress or strain distribution through the entire wall thickness.” That’s easier to think about, and quantify.
What Causes Discontinuity Stresses?
A few examples might better illustrate the source of discontinuity stresses.
Example #1
Major and minor (a.k.a., local) discontinuities are related to the stress intensification that occurs at the discontinuity. These stress intensifications are the result of a discontinuity in the stiffness of the structure. Consider the simple example of a cylindrical vessel with a change in wall thickness (Figure 1).
On pressurization, the thicker and thinner portions of the cylinder would naturally experience different amounts of radial growth. In this example, the ½ inch thick cylinder would grow 0.028 inches (radially), but the 1 inch thick cylinder would grow only 0.014 inches (both shown in Figure 1B). But if they are welded together, they must grow the same amount at the weld line. That “compatibility condition” results in bending in both cylinders, as seen in Figure 1C.
The resulting bending stresses are exactly what they are talking about in Section III, Division 1, paragraph NB-3213.2 when they refer to through-wall bending.
Example #2
Another example of discontinuity stresses at a major structural discontinuity can be seen at a cone-to-shell junction. Figure 2A shows a typical cone section of a vessel. Figure 2B shows the corresponding longitudinal stress distribution.
Figure 2B shows significant through-wall stress distribution extending a significant distance from the corner.
But let’s look at the stress distribution in more detail. Figure 3 is a plot of the longitudinal stress on the ID (blue line) and OD (red line), starting at the corner (“Point A”) and moving upwards. Comparing Figure 2 and Figure 3 yields some interesting insights.
- Both figures show that at the corner, the ID starts out in tension and the OD is in compression. Moving up, away from the corner, the stress concentration lessens as the stress approaches the nominal longitudinal stress.
- Figure 2 shows that the through-wall stress variation seems to have pretty much evened out by 6t away from the corner (i.e., at “Point B”). But the plot of Figure 3 extends farther and shows that the stress really doesn’t level-out until about 30t from the corner!
- Figure 3 shows a bit of oscillation in the stress pattern as the OD and ID both over-shoot the nominal longitudinal stress (pr/2t) before they settle down around 30t from the corner.
- There is a rule-of-thumb often seen in API 579, and other places, that a separation of 1.8√Dt is required to prevent interaction of local disturbances. This rule-of-thumb serves reasonably well in this case. Figure 3 shows the stress perturbation is fairly well attenuated at 1.8√Dt. However, it’s not completely attenuated until a bit past 2.0√Dt, or about 30t in this case.
Lastly, Figure 4 shows the through-wall bending stress from the ID to the OD, and at different distances from the corner. Note that the bending stress is greatest at the corner, indicated by the steepest slope on the blue line and the highest stress at “Point A”. At 6t from the corner (horizontal, purple line), the stress distribution is flat, with no bending.
So, remember that when you hear the term “Major Structural Discontinuity” or “MSD,” you need to think of two things:
- Through-wall bending stress, and
- Stress occurring over a relatively large area (or volume) of the structure.
The term “relatively large” is not very quantitative, but it would probably be fair to say that a stress concentration that dissipated within 1t or 2t of the source, or within, say, 0.25√Dt to 0.5√Dt, would probably not reasonably qualify as an MSD.
List of MSDs
All of the documents cited in the above definitions provide examples of typical MSDs in pressure vessels. Thus, compilation of a fairly comprehensive list of MSDs is pretty easy. In Table 1, I’ve compiled such a list of MSDs and LSDs given in the most common references.
| Major Structural Discontinuities | Local Structural Discontinuities |
| Head-to-shell junctionFlange-to-shell junctionNozzles-to-shell junctionJunction between shells of different diameter or thicknessSupport locationsVertical vessels – skirt, leg, or base junctionHorizontal vessels – saddle supportSphere – Leg junctionTank – Floor-to-wall intersectionLifting lugsSupport Lug connection (e.g. pipe support lug)Stiffener ringConical transitionTank bottom shell-to-chime plate junction | Full penetration structural welds used in vessel constructionLong seam, circumferential seamHead-to-shell weldSmall attachmentsSmall fillet radiiPartial penetration weldsOil hole, or weep holeMerger of knuckle and crown radiiSmall grooveWeld toe |
One last point – Even among the MSDs listed in Table 1, there can be still some ambiguity. Consider a support lug supporting piping. It could be a small lug holding a 1 inch pipe, or a large, heavily-loaded lug supporting a 24 inch pipe. Technically, whether it qualifies as an MSD or not depends on whether the stress distribution is through-wall and the extent of the stress perturbation. Of course, a finite element model could be used to help decide if this should be treated as an MSD or not, but once you’re building a finite element model you’re really leaving the Level 2 analysis and moving into the realm of a Level 3 analysis. So, in practice, the list of MSDs in Table 1 may be conservative, but once you go down the path of evaluating the severity of the MSD, you’re probably in Level 3 territory anyway.
Closure
Understanding what is and is not a major structural discontinuity is essential for most Level 1 and 2 FFS assessments under API 579-1. Hopefully, this article has helped you get a better understanding of MSDs and will add to your comfort level in performance of FFS assessments.
If there are any topics you’d like to see in the FFS Forum, or if you have comments on this article, please send me an email at ffs@inspectioneering.com.
Thanks, GG
P.S. Special thanks to my colleague, Jack Hawkins, P.E., in the New Orleans office of Stress Engineering Services. Jack built and ran the finite element model used in this article.
Original post by: Greg Garic