RE: Liberal Allowable

From: <Tony>
Date: Tue Aug 23 2005 - 14:09:00 EDT

Re: Use of Liberal Allowable

As one of the original developers of the CAESAR II program, I personally have made the following statements regarding the liberal allowable in various seminars:

1. The "liberal allowable" as used in CAESAR best matches the recommendations of ASME Section VIII Divison 2 Appendix 5 (see B31.3 304.7.2), and ASME Section III Subsection NB3200 for fatigue stress calculations in intent. Almost any other adjective would have been better than "liberal." My personal preference is "Accurate Allowable."
2. In all cases, where extra conservatism is justified, the designer is urged to apply that conservatism as he sees fit. This might include, omitting the so-called "liberal" component of the allowable, not permitting stresses to approach allowables within a restricted range, etc. Situations where this might be considered include: Pilot plant construction where operation may be erradic, or a any occasion where the described loads or corrosion rates are unknown or difficult to predict.
3. The term "liberal" was a misnomer from the first time we were aware of its use in 1975 (or thereabouts) by one of the first mainframe programs to be converted from Mec 21 (the "Mare Island Program") to a commercially viable form. It is a shame that this name was selected as it implies an desire to provide safety which is needed, and we do not think that this was ever the intent. The name has no political or "real" implications that we know of except that for stress analysis it is commonly felt that any calculation or approach which permits a higher stress is more liberal than one that doesn't. Pump vendors, for example, that permit zero loads on their pump nozzles, are conservative and not crazy,... er I mean liberal.

-----Original Message-----
From: PipingDesign@yahoogroups.com [mailto:PipingDesign@yahoogroups.com] On Behalf Of Christopher Wright
Sent: Tuesday, August 23, 2005 11:13 AM
To: PipingDesign@yahoogroups.com
Subject: Re: [PipingDesign] (unknown)

On Aug 23, 2005, at 12:23 AM, Pankaj Mandal wrote:

> Erik is right. B31.1, Para 102.3.2 (D) permits the use of liberal
> stress for piping calculations. In simple terms liberal stress allows
> use of unutilized stress margin in sustained case as an additional
> stress allowable for expansion case.

Be a little careful about this. Codes recognize higher allowables for certain combinations of primary + secondary stresses only. Be careful not to go using your 'liberal' allowables with primary calculated stress. Same thing holds for fatigue analysis.

Primary stresses result from statically determinate loads, like pipe walls under internal or external pressure. When a primary stress exceeds the yield stress, the pipe will deform without limit. Secondary stresses are self-limiting and include discontinuity stress and certain thermal stresses, where the load is statically indeterminate. An example is the discontinuity stress that develops between the head and the shell of a pressure vessel. In that case yielding tends to redistribute the stress without causing plastic collapse.

Bolted flange calculations have similar provisions for hub stresses, which involve secondary (discontinuity) stress. The remainder of the calculated flange stresses are primary stresses and the primary stress allowables apply. You don't want to mix them up.

Christopher Wright P.E. |"They couldn't hit an elephant at chrisw@skypoint.com | this distance" (last words of Gen.

.......................................| John Sedgwick, Spotsylvania

1864)
http://www.skypoint.com/~chrisw/

---

PipingOffice - Excel Spreadsheets for Piping Calculations http://www.pipingoffice.us/

---

Main site: http://www.pipingdesign.com

Yahoo! Groups Links Received on Tue Aug 23 14:09:00 2005