Maximum circumferential stress in highway crossings in pipelines 2

[dasumit]

Hi,
I am a newcomer to the pipeline engineering field and want to get a clarification regarding a doubt I had while calculating Maximum circumferential stress in highway crossings. Any help will be really appreciated.
According to API RP 1102, while we perform the Principal Stress check, while calculating maximum circumferential stress, we add the stresses due to Earth load, highway cyclic stress and the stress due to internal pressure. My question is that , why do we add all three even though stress due to internal pressure acts in an opposite direction to the other two stresses? Like the first two stresses are due to external pressure on pipeline while the other is due to internal pressure in pipeline .

Thanks
Sumit Das.

 

[BigInch]

Exterior loads, unless from a uniformly applied hydrostatic pressure, do not produce ring stresses. Surface loads are applied to the top and the bottom of a buried pipe with a resultant that could be better described by point loads on top and bottom. Those point loads produce bending stresses in the walls. The tension developed by those bending stresses adds to the hoop stress, even though the compression subtracts from hoop stress. The critical stress in that case is the sum of all tension forces!

What would you be doing, if you knew that you could not fail?

 

[dasumit]

Hi BigInch,
Thanks for the reply. But I am confused that in this situation, the bending will contribute to longitudinal stress and not the circumferential stress. Because in general design of pipeline as well , we consider stresses due to bending moment as a contributing factor in longitudinal stresses, not in circumferential stress. Whereas in this case, according to your explanation , they seem to be adding the bending stresses in calculation of maximum circumferential stresses instead. Why is there such an anomaly.

Thanks for your help,
Sumit

 

[dasumit]

Hi BigInch,
Thanks for the reply. But I am confused that in this situation, the bending will contribute to longitudinal stress and not the circumferential stress. Because in general design of pipeline as well , we consider stresses due to bending moment as a contributing factor in longitudinal stresses, not in circumferential stress. Whereas in this case, according to your explanation , they seem to be adding the bending stresses in calculation of maximum circumferential stresses instead. Why is there such an anomaly?

Thanks for your help,
Sumit

 

[BigInch]

Longitudinal axis bending what you're thinking about, is for beams. An underground pipeline is "theoretically" continuously supported, so there would not be any of that. You would especially not want unsupported spans under a highway, right. A point load applied to a "wedding" ring creates bending "parallel" to the hoop stress in the wall (see attachment, hoop stress not shown), just as in an arch. Check out stress analysis of an arch. A pipe is actually a ring in cross section (two arches attached together where the bases can move apart if stresses get high), and only if the pipe is long enough and unsupportted, also a beam. If longitudinal beam bending stress is present, you must then add the transverse tension side ring bending stress to the hoop stress and then combine that with the longitudinal beam bending tension stress at the "bottom" of the "pipe beam". OK, that was for tension allowables. Now consider the addition of all compression side stresses for local buckling failure and compression collapse in the transverse ring structure, as that is what you're really interested when checking surface loads. The longitudinal stresses just add to the problem.

What would you be doing, if you knew that you could not fail?

 

[dasumit]

Thanks BigInch for a very instructive answer.
Sumit Das

 

[rconner]

In a period roughly 50-60 years ago, and coincident with tremendous expansion of both transportation and pipeline infrastructure in the USA, there was great interest in pipeline crossings (where such pretty much inevitably intersect). One of the charter technical activity committees of the ASCE Pipeline Division, i.e. “Committee on Pipeline Crossings of Railroads and Highways”(and also a collaborative “Research Council ), were notable groups charged with examining the effects of loads on and designs of pipeline crossings. These folks looked primarily at common crossings of highway and railroads, but I think also included other sorts of traveled crossings (such as airport runways etc.) At the time while most such crossings of the day (as I believe is the case also now) were combined casing and carrier installations installed by either open-cut or various means of jack and bore, there was also knowledge at the time of some uncased crossings (particularly with strong pipes such as steel and ductile iron etc.) If you are keen on examining this subject in detail, it might help to look over/stand on the shoulders of these piping giants.
In January of 1964, this Committee published a nearly 40 page technical report in ASCE Proceedings Volume 90 NO. PL1 Journal of the Pipeline Division, entitled ”Strain-Gauge Analysis for Uncased Pipeline Crossings”. This report documented detailed pressure and load testing in simulated highway and railroad embankments at the Association of American Railways Laboratories, employing 12” steel pipe with about ¼” wall thickness at about 2-1/2 feet of cover surcharged with jack-simulated 10-15 kip wheel loading (with steel plate centered over pipe on the surface) as well as jack-simulated railroad loadings up to 95 kip (with a three-tie and rail arrangement with the pipe centered under the middle tie and jack load resembling individual engine axle? on the rails above). Cyclic loading (up to 2,000,000 cycles) and internal pressures up to 1,000 psi were included in some tests. In general I believe stresses encountered due to external loads were low, and railroad stress results only some higher than highway. In June of 1964, M.G. Spangler (who also consulted on the aforementioned work) referred to same and also presented more results of extensive research by the Council involving “experimental” and many actual and larger diameter pipe/casing crossings around the country at the annual AWWA Conference in Toronto, Canada. This paper was published in the AWWA Journal paper, “Pipeline Crossings under Railroads and Highways”. Professor Spangler of Iowa State University (a few years later my alma mater) was incidentally part of a long line of notable pipeline researchers at that institution going back now more than 100 years. These included Anson Marston, W.J. Schlick, and just a little more recently though now also well-known student Reynold Watkins in Spangler’s time. This Spangler paper provided research results involving most common unpressurized steel etc. casings actually installed beneath working railways with locomotives of from 256,000-336,000 pounds and speeds of from 33-100 mph (as well as stopped over the pipe). It also talked at length of the effects of other loadings, pressure and “re-rounding” effects etc. of direct buried pipes (referencing a few months earlier published ASME and AWWA papers by E.C. Sears of AMERICAN, regarding trench testing of similar large-sized direct buried and pressurized ductile iron pipe). This Spangler paper looked at a great many aspects and concerns involving particularly typical bore and jacked casings, including strains/stresses (where possible as determined w/ strain gauges), actual deflections, both static and dynamic (with measurements at the same center-punched measurement locations on the pipes etc.), the extent to/manner in which soil sort of in many cases moves back into contact with the pipe at different clock locations (e.g. from a typical 1-2” diametrical tunneling/boring over-cut), “vertical sag”, empirical determination of effective/developed E’s, and also an interesting observation of varying invert elevations of several abandoned casings that had been installed for many years. These latter were not installed in the research program and therefore initial installation conditions/data were not known, but the casings were in otherwise good condition and they could now be examined in detail. While very little ring deformation was evident, all these several year-old crossings exhibited a phenomenon called “crowning”, where the inverts near the middle of the crossings were found to be from 3-11 inches higher at the transverse centers of the highways involved, than they were at the ends. While I’m not sure based on otherwise lack of comment concerning this in the paper that even the brilliant Spangler at least at the time fully understood the latter find (that I suspect may not be real surprising to some contemporary folks experienced with long-term piping through substantial embankments, as many roadways sort of resemble), in any case he could not conclude anything as he did not have initial installation measurements/data etc. Spangler likewise voiced little else in the way of loading concerns as a result of the detailed examinations of the 30”-42”, ¼”-3/8” respectively thick researched steel casings, where substantial before and after data was available. This appeared to pretty much explain a comment made earlier on in the “Field Studies” area of his paper, “Although literally thousands of flexible casings are in service under railroads and highways and no one has ever documented a failure or an instance of excessive structural distress of such pipe, nevertheless, specific knowledge of their service performance was nonexistent.”