According to page 6-26 of the user manual
for version 4.5:
"if the user has an operating temperature below
ambient in addition to one above ambient the user should add another load case as follows:

Case#6 L2-L1 (Exp)

(assuming t1=hot t2=cold)

Am i right in thinking that this effectively
an addition of the cold stress to the hot stress?
ie a Subtraction of a negative stress is an addition?

If this is so, is it correct as the pipe will pass from the cold case (t2) through its relaxed (ambient) state to reach the hot case t1?

Also, is this a code specific requirement as
B31.3 does not require the addition of stress's
that pass through ambient.

I am particulally intrested in this as i am often
doing calculations that pass from cold to ambient to hot.
and i assume the pipe is relaxed as it passes through the ambient stage.

This topic was discussed in an article in the July 2003 edition of COADE's Mechanical Engineering News. The direct link to this newsletter is here.

Read the article Satisfying Expansion Load Case Requirements, beginning on page 11.

One thing I'd like to add...

A new configuration switch was added with the release of Version 4.50 under SIF's and Stresses - Press. Variation in EXP Case. This switch has three settings - Default, Yes and No.

You might have noticed that our recommended loads cases always used P1 in the operating and installed cases. For example, with W, T1, T2, P1 & P2 defined, the following load cases were recommended:
1)W+T1+P1 (OPE)
2)W+T2+P1 (OPE)
3)W+P1 (SUS)
4)L1-L3 (EXP)
5)L2-L3 (EXP)
Note that P1 is in both Case 1 and Case 2 and so there is no delta P showing up in the expansion stresses.

So, if you change P1 to P2 in Case 2 and add Case 6, your load cases now look like:
1)W+T1+P1 (OPE)
2)W+T2+P2 (OPE)
3)W+P1 (SUS)
4)L1-L3 (EXP)
5)L2-L3 (EXP)
6)L1-L2 (EXP)

Prior to Version 4.50, the change in logitudinal pressure stress would appear in Cases 5 and 6. Now, depending on this new switch setting, the change in longitudinal pressure stress (P1-P2) may or may not be included in Cases 5 and 6.

I suggest you decide how you want to handle this issue and change the switch to Yes or No accordingly. There are proponents for each approach. I'm just telling you why your numbers may change.

Your assumption is very much incorrect as far as B31.3 is concerned.

You'd best take a look at B31.3 302.3.5d, 319.2.3 and 319.3.1. The stress range is from minimum to maximum metal temperature. For most systems, it is common to take the ambient "as installed" condition as the bottom of the range. However, if you've got cases where you system is operating both above and below the install temeprature, your total displacement stress range has to cover that full temperature range.

That's exactly why the extra case above is recommended.

Dear sir,

The previously stated paragraph "The stress range is from minimum to maximum metal temperature"

is taken from:
ASME B31.3 section 319.3.1(a)

However, This paragraph has a condition:
that is " for the thermal cycle under analysis"

Is it not the case that as we pass through the installed state or 'ambient' Temp, we are then looking at a seperate cycle?

To clarify, (say)when passing from very cold
Through ambient to very hot, we are in fact looking at two different thermal cycles as all stresses are normalised at ambient?

With respect to para :319.2.3(b)

My interpretation of this is for example,where a pipeline may have been installed at a different temperature than ambient (in winter for example)

Your thoughts on these matters will be apreciated.


That is my interpretation of the code.
Has this issue been clarified by ASME??

B31.3 para. 319.2.3(a) focuses on a single cycle cycle to describe the effects of yield or creep - conditions that may cause self-springing at the completion of "the cycle". Para. 319.2.3(b) makes it clear, though, that it's the maximum displacement stress range between any two conditions that's used to set SE - the computed stress range.