The thermal stresses are considered secondary as they are disp based stresses and disappear or get relaxed on local yielding. But the Weight remains always and they dont get relaxed or disapper, so they are primary stresses. The stresses due to seismic are also considered primary, but my question is won't they get relaxed or disappear by local yielding as they act only once or for a very small period. please clarify.

Well, yes and no.

Current piping codes view all force-based loads as primary where failure by collapse is the concern. (Secondary or displacement-based loads focus on fatigue failure.) For seismic load, the souorce of the load is inertial - all supports move with the ground or structure and the pipe has to "catch up" with the supports. To prevent failure, the pipe is not allowed to yield.

However, the yet-to-be-released B31E (which mimics the American Lifelines Alliance document on seismic design of piping systems) uses a fatigue approach to set the limit on stress. A certain number of cycles are assumed for the duration of the seismic event and from that, an allowable stress is determined. It's a rough calculation (in a document that uses more design by rule than design by analysis) but it is a shift away from primary stress evaluation of seismic loads. Further to your point, as I recall, gaps of up to 2 inches on guides can be considered as active supports in the seismic analysis. Even though these gaps my be open in the operating state, the contact made during the seismic event will effectively "kill" that high modal response. To account for that imnpact load, the support must be designed to carry two times any static equivalent load calculated at that support (DLF=2).

I believe we are in a transition in the evaluation of seismic loads.

the line has two different loads to accomadate. First the collapsing load which is adressed by "occasional" in the current code.

If the line does not collapse then it is subject to low cyle fatiuge as it is shaken back and forth. Currently the codes do specifcally adress low cycle fatiuge.

Hello rajagopal,

Thank you, we enjoyed your topic of discussion. You are obviously thinking and that is commendable.

Your comment that "secondary stresses ..... disappear or get relaxed on local yielding", compels some additional comment. Thermal expansion driven displacements may cause local yielding of the piping system (local areas of plastic deformation - usually in bends and elbows) that, over several cycles, often "shakes down" (i.e., "relaxes") to purely elastic response due to local strain hardening and the redistribution of strains. When fully "shaken down", the piping system's response to subsequent thermal cycles will be purely elastic response. That is a good thing and the fatigue rules of the B31 Piping Codes are based upon it (the allowable stress range is set accordingly). Of course, fatigue is an issue over an extended period. If a local “stress riser” (e.g., a discontinuity – geometric or metallurgical) facilitates the development of a fatigue crack in the pipe wall after a period of thermal cycling, the crack may propagate through the pipe wall and cause a failure. The B31 Code rules provide protection against fatigue (the rules are based upon the Markl cyclic load tests). If the B31 thermal displacement stress rules are followed and the allowable stress range for the material at temperatures is respected, the system WILL "shake down" over succesive cycles and fatigue will not occur due to thermal cycles. However, if the pipe does not fail by fatigue induced cracking, and if the hot and cold yielding of the piping material, over many thermal cycles, is too great (on successive alternating load applications) and complete "shake down" cannot occur, the system may fail due to accumulative "ratcheting". But fear not, for this to happen, the same area on the pipe wall would have to see unrelenting yielding at every hot-to-cold and cold-to-hot cycle. The stresses would have to be well beyond those allowed by the Code for thermal displacement (your CAESAR II Code Compliance Report would “be waving a red flag at you” assuming the proper thermal loadings are anticipated and evaluated). However, thermal expansion displacements are not the only alternating loadings that piping systems may encounter. Alternating force induced displacement stresses high enough to cause unrelenting alternating yielding are possible due to other (other than thermal cycling) types of loadings. If the piping system is subjected to alternating loadings of sufficient magnitude and relatively short duration (e.g., ocean wave movement during storms or high wind induced movements of tall stacks (Aruba comes to mind) OR SEVERE SEISMIC EVENTS) and the yielding pipe does not form a plastic hinge, a "ratcheting" failure cannot be discounted. So what?

Full scale testing of "typical configuration", three-dimensional piping systems was done (on a very large "shake table") several years ago. Shaking profiles representative of large magnitude three-dimensional seismic events were applied to the full-scale systems and the results were interesting. When failures in the piping occurred, the failure mechanism was accumulative ratcheting due to alternating excursions into the plastic regime - not collapse due to "single application in one direction". The load direction reversed before a plastic hinge could form in any unidirectional load application. So, the immediate question was "is this 'primary' or 'secondary' in nature". As the testing was done at the behest of a certain nuclear power regulatory authority, the next question was what SIF's should be applied in the calculation of such stresses - the primary or the secondary load SIF's? Or, maybe a better question is “do we need ‘Occasional Load’ SIF’s for B31? We don’t have “Markl-type” data for loadings of this type. So, we still have to get this issue sorted out and sell the regulatory authorities on the obvious (to us) correct concept and then appropriate rules have to be written and voted into the Codes. The B31 Seismic guideline document has been in development for over 20 years. Maybe the best question is "will I live long enough to see it"?


Regards, John

John your comment... ""will I live long enough to see it"?"

Maybe.... the good die young, you may have a long life ahead of you!!!! LOL

Rajagopal,

Thermal stresses are not always secondary, then can be peak also. Read Appendix 4 of sec VIII Div 2 for detail.

Also there are situations where thermal stress can actually act as a primary stress.As an example,if a thin tube is subjected to a length-wise temperature profile that has a sharp jump discontinuity at some point along the tube and this discontinuity moves repeatedly along a length larger than, typically, a couple of radii of the tube, then the shakedown limit is near equal to the elastic limit, i.e. 2 is replaced by 1.
For details on this subject , you can type " A.R.S.Ponter" in google, Professor A.R.S.Ponter is an international expert in this field and you will get more information.

Regards

Ahh yes Anindya is 110% correct also google elastic follow up....

The shake table John Breen alludes to was a piping set up that I felt was unrealalistic... as I recall it was a rather long L bend with a valve on one end of the L and the other end "anchored"

When the shake table shook... the piping did not immediately collapse the concern of the occasional loads in the codes now, but after ~ 100 cycles or so it cracked. This could be described as low cycle high stress fatigue.

..........and not all displacement stresses are caused by temperature changes. Not all "thermal stresses" are cyclic thermal displacement stresses. And not all possible stresses are explicitly addressed by the B31 Codes. Watch B31.1 in its next edition for changes in some terminology and the introduction of a definition for "peak stress".

Regards, John.

yeah well I still don't like adding a term to a code which does not use the term... and so guess where that one will go in B31.3!

If an indidual is ambityous enough they will pursue matters on their own and pick "peak stresses" along the way.

For crying out loud nobody reads the damned books anyway!

Grrrr

Thank you everyone for your answers.