I am analysing a DN600 Class 900 line with very long lengths to B31.3 with V4.50, and would like to take into consideration pressure effects. I can see that there are 4 field options:
In configuration file setup:
Pressure stiffening effects: Default/Yes/No (for B31.3 Default=Yes), uses max pressure defined, I assume the defined Hydrotest Pressure in the input file is not used here.
Bourdon Pressure Effect: None (default)/Trans/Trans & Rot
In Special Execution Parameters:
Activate Bourdon Effects:None (default)/Trans/Trans & Rot
Stress Stiffening due to pressure: None (default)/Press#1/Press#2

Why have 2 Bourdon effect fields? I get different stress results for each of the following:
1. Config file: None
Exec Para: None

2. Config file: None or Trans
Exec Para: Trans

3. Config file: Trans
Exec Para: None

It seems that the exec para field overrides the config file setup field (per 2 above) for Bourdon effects, but 3 gives a different result to 1, why is this?

I presume the fields should be set as below to fully consider pressure effects:

In configuration file setup:
Pressure stiffening effects: Default or Yes (for B31.3)
Bourdon Pressure Effect: Trans (line has welded elbows)

In Special Execution Parameters:
Activate Bourdon Effects:Trans
Stress Stiffening due to pressure: Press#1 (my design or max operating pressure)
Regards
Mike

Pressure Stiffening Effects: This Configuration Option controls whether or not CAESAR II implements the pressure stiffening effects on bends when computing the SIF and flexibility factor. My advice here is to leave this directive set to DEFAULT which means do what ever the active Piping Code says to do. So if the Code implements pressure stiffening effects on bends, then so will CAESAR II. If the Code does not implement pressure stiffening, then neither will CAESAR II. The other two settings (Yes and No) are there so you can force the issue one way or the other (say for benchmarking purposes).

Hydrotest Pressure: This input value is only used when the Stress Type of the Load Case is set to "HYD".

Bourdon Pressure Effects: Read the Help Text on this option from the Special Execution Options Dialog. It should explain the differences in the two settings (Trans and Trans+Rotation).

Stress Stiffening due to Pressure: I strongly suggest you NEVER USE THIS OPTION. Sometime back a user presented an argument for including stress stiffening of straight pipes due to internal pressure. The mathematical argument (60+ pages or so) made sense at the time. We then made a few ANSYS runs to prove the theory, and convinced, we added this abiltiy to CAESAR II. However, since then, additional usage and testing of this option has produced suspicious results that we can't justify. For Version 5.00, this option has been removed from the software.

You are correct, the settings on the Special Execution Options Dialog over-ride those from the Configuration File. The same holds true for the ambient temperature value and the switch for the liberal (expansion) allowable. The idea here is that the Configuration File applies Directory wide settings that are used to seed new input files. However, once you're in the input, you're free to change these values. This options are provided simply as a baseline starting point and as a time saver.

Mike,

You have hit on what in my mind is something that B31.3 may have the wrong (??) or perhaps (misleading) take on, in some certain circumstances....

First look at B31.3 Appendix D (note7) and you will see the pressure stiffening equation listed. This stiffening occurs in the elbow and along with stiffening the elbow it reduces the SIF of the elbow. So the disconnect that may occur is the relationship between P1 and T1.... if T1 produces the greatest thermal strain and if the internal pressure that is present during T1 is actually lower than the internal pressure max to use the higher pressure coincident with the higher temperature would in a sense be non-conservative and yield an incorrect answer! (Usually the percentage error would be small)

The bourdon effects are not code requirements but are used to simulate the straightening effect of the elbow to pressure a similar effect acts on "Bourdon" pressure gages. (Think of the small child’s toy used at parties when blown into the toy straightens out from its curled up shape)

So what to do???? I would leave out the Bourdon effect(s) and use the code required pressure stiffening but I would try to use the appropriate pressure.

However one other little wrinkle... the internal pressure term is used to calculate the sustained load stresses.... so the pressure should not be underestimated either because as internal pressure goes up it will increase the sustained load stresses!

Confused??? Me too, the thing to keep in mind is that ordinarily these effects are minor and of no major consequence.

Richard/John,
Thanks for your prompt replies and expert advice. It is now clear what to use.
As you mentioned John, there are some load cases that may not be as conservative as at first envisaged. My line is along an offshore pipe bridge between 2 platforms. Using just the default pressure stiffening effect as advised above, my worst case displacement stress range code check (which occurs on an elbow)is when using the design pressure, as when using the liberal allowable this gives me a lower allowable stress. However, when using the operating pressure, the calculated max stress range is higher (due to higher SIF of elbow as you said), but the liberal allowable stress is also higher, giving me a lower % of code allowable. The operating pressure case thus gives me a higher stress range with which to carry out my fatigue assessment (using EN 13445-3 Section 18). In fact during a 100 year cyclonic storm, the process should be shut in with no pressure, and the max stress range is even higher but again the allowable is higher too, giving a lower % of code allowable, then when using the design pressure. The difference in my case between using design pressure and zero pressure, is 29 MPa (4 ksi) which is quite significant, but as you said will generally not be significant.

Richard,

Just for interest, why do you get different stress results between having the default "none" specified in both the configuration setup file and the execution parameters for Bourdon effects, and with having "trans" in the configuration file and "none" in the execution parameters for Bourdon effects? As you confirmed, the execution parameters should override the configuration setup file, and you would expect the same result.

Best regards
Mike
PS I liked the attachment you provided John, sure brightens up this forum and in the festive spirit too, talking of which I must dash to my agent's xmas party.

You shouldn't. However, you need to be careful when changing configuration directives that affect the computations. The only way for the changes to get into the analysis is to go back through Input and Error Checking. If the software worked any other way, then changing a configuration directive could affect every other job in the directory, even if you didn't want it to.

So my first thought is that the input was created with "trans" in the setup, then after the run the configuration and the input were changed. (Note that this particular directive only affects "new" jobs.)