I suppose this topic has been discussed many times in the past (found some relevant discussions) - yet posting it for expert opinion and clear understanding of the subject.

For a piping system being designed as per B31.1, we have following typical temperature conditions:
Average Ambient (expected condition during installation) : 24ºC
Minimum ambient : -4ºC
Design temperature for the piping system: 220ºC

The prevalent practice in the organization I have recently started with is to analyze the model by defining the ambient as 24ºC and operating temperature (T1) as 220ºC. The minimum ambient temperature was not used at all in the analysis.

Considering the fact that there is theoretical possibility of the plant shutdown during the coldest ambient and the metal temperature reaching the minimum ambient (or a nominated temperature, close to the minimum surrounnding), I believe the above philosophy is wrong.

To check for the correct full thermal stress range, I feel we should do one of the followings:
i) Define ambient as the minimum site temperature, ie, -4ºC. The only downside I can think of is that the physical thermal expansion (movement) will be calculated for -4º to 220ºC which is not correct.
ii) Define another operating condition for T2 = -4ºC @ 0 barg pressure, which will enable checking the thermal stress range for T1-T2

Would appreciate suggestions from experts as to what is followed in other organizations for this common situation.

regards,

hi sanjay,

I am not an expert but I incounter in my reading somewhere in code interpretation, guided by last question I post in this forum.. That even in the occurance of the temperature in 5seconds that shall be evaluated for thermal stress range provided it is the max or the min temperature for its whole life..

Regards!

According to the codes, which speak about "stress range" you should take the stresses over the complete range from -4 to +220 degr. C.
Personally I would use two cases with T-ambient at 24 degr C. That way you'll have the displacement figures correct.

The method used in your company, checking from 24 to 220 degr C. only, is used in a lot of other companies too. Here in The Netherlands I always see that done, instead of checking the complete stress range.

Dear Sanjay,
The situation you are describing is common in our industry.
The point is that the displacement range that we are talking about is the range between the highest temp (design in your case) and the lowest temp (minimum ambient in your case).
Please read the july03 coade newsletter to get the better understanding on the subject.

Corne,

Not all companies in the Netherlands neglect the total thermal stress range (e.g. the company where I am working).

I would suggest to make an OPE case with T1= -4 deg. and one with T2= 220 deg. Then you can make an EXP case which extracts those two OPE load cases, thus the total thermal stress range.

Regards from The Netherlands,
Jeroen

Jeroen,

I didn't say all companies don't check it correctly. Only thing is that the stress-analysis I see from other companies do it incorrectly. Glad to hear there are more people doing it correctly. May I ask where you work?

Caesar Forum Members,

I am pretty much confused with the above discussion, i dont find any ideal solution in the above discussed "Thermal Stress Range" topic. Some of our member says, we can find the thermal expansion stress range between "Design temp (T1) (Maximum temperature) and Mean ambient temp (T2) (Minimum temperature)" and some of them says it is the diff between "Design temp (T1) (Maximum temperature) and Minimum ambient temp (T3) (Minimum temperature). From my point of view, we use the latter one i.e (T1-T3) as per ASME B31.3.

Normally, for medium and high temperature service, the thermal stress range for the latter one i.e (T1-T3) will be more than the earlier said option (T1-T2). If there is that much controversy between many companies of the world, then definitely some companies are doing some errors in their Stress Analysis. Also our friend Mr. Rajinder Singh asked us to refer the Coade newsletter - july 03, from that, my understanding is to use (T1-T2), i just want Mr. Rajinder Singh to confirm whether my understanding was rignt or wrong.

I request the experienced members to give the perfect solution regarding the thermal stress range. So that in future these kind of errors/confusion shall be eliminated.

Don't get too concerned by all the T1-T2 stuff. That is for the user to sort out.

B31.3 is quite clear on this.

For example, in a typical high temperature system, installed at a moderate temperature but subjected to a low winter temperature, the stress range will be the algebraic sum of the installation to high temperature stress and installation to low temperature stress. Other cases will be needed for equipment and support loads etc.

If the installation to low temperature part of the stress range has been ignored, the calculation is simply wrong. Maybe not by much, but still wrong.

........so in summation:

For use with the B31 Code for Pressure Piping:

The full thermal range is the coldest temperature that the piping will experience to the hottest temperature that the piping will experience. It would be the summation of the installation temperature to coldest temperature and the installation temperature to the hottest temperature. Keep in mind there might only be a few full cycles and many, many partial cycles and that is why the B31 Codes provide an equation for calculating the equivalent total cycles (and this becomes the indicator for the "f" value in calculating the allowable stress range.

Also remember that the B31 Codes currently require the use of the materials modulus of elasticity at 70 degrees F when calculating the thermal STRESS range for these thermal (displacement) ranges.

When you calculate the loadings (forces and moments) on the equipment you would use the OPERATING temperature and the modulus of elasticity at the operating temperature. You may find that this makes a great difference in very hot systems and very cold systems.

You can see why it might be necessary to make quite a few loading cases in your analyses.

Regards, John

Dear paldex,
Let us say that the piping has operating temp T1, design temp T2 and minimum ambient temp T3.
The piping will be installed at the installation temp. When the piping will become hot it will see the operating temp T1 and if temp continues to rise it may see temp T2. Similarly the piping will also see T3 if it is cooled. In order to represent the above conditions we require the following load cases.

1) W+T1+P1 (OPE)
2) W+T2+P1 (OPE)
3) W+T3+P1 (OPE)
4) W+P1 (SUS)
5) L1-L4 (EXP)
6) L2-L4 (EXP)
7) L3-L4 (EXP)

The load cases 5), 6) and 7) take care of the three conditions mentioned above. But are these three load cases sufficient enough to satisfy the intent of the code regarding displacement stress range? Answer is No.
To completely satisfy the intent of the code, another load case must be added, as follows:

8) T2-T3 (EXP)

The above load case will take care of the condition that will cause the highest differential effect as mentioned in B31.3 para 319.2.3.b.
Hope it clarifies.

RAJINDER IS RIGHT

...and please refer also to the link below:
http://www.coade.com/ubbthreads/ubbthreads.php?ubb=showflat&Number=19663&page=5#Post19663

...and the pdf file at the bottom of page 2

All,

Thanks all for giving valuable suggestions and interpretations. This is the beauty of such forum to clarify such doubts by seeking opinion and through discussions.

As a new user of Caesar, will appreciate somebody explaining an elementary thing I noticed in actual analysis, as described below.
In line with the recommendation of this discussion, for the particular piping system I'd defined T1 (design temp) = 220ºC, T2(min operating)= 0ºC and ambient (in configuration set-up)= 22ºC.

I got a warning that "... the temperature for Case 2 is entered as 0ºC. This will be assumed AMBIENT.." The basic load cases did not include T2 separately either.

While this will give the true expansion stress range as we discussed, I expected the ambient will still be 22ºC so that I get true thermal movements for installtion-design temperature case. It seems Caesar automatically assumes the minimum temperature as AMBIENT. If that is the case, then instead of defininf another temperature case, we should set the ambient value as the minimum temperature (instead of average or installation temp).

I know I am missing something (due to lack of experience with Caesar) -- would be good if somebody can point that out.

Thanks in advance,

Avoid using 0.00 as a temperature value. Use something just larger than the "alpha tolerance", otherwise the value you input will be treated as a "strain" (hence them message regarding 0 being ambient).

By default the "alpha tolerance" is 0.05, so any value you enter less than this will be taken as a strain (or the expansion coefficient in units of length/length).

So instead of T2=0.0, use:
a) 0.051, and if this isn't sufficient then
b) In the Configuration Module reduce the "alpha tolerance" value, to say perhaps 0.001 and then define T2 as 0.002.

Read the "Help Text" on the temperature field and the "alpha tolerance" field for additional insight into this.

Great.

I had a read of of the Help Text and realized what was happening - have fixed the problem.

Thanks Richard

Just two more points on expansion stress range...

1) There's an example of "hot" and "cold" states in B31.3 Appendix S. There the ambient-to-hot stress range is added to the ambient-to-cold range to get the total stress range. I get the same numbers running cold-to-hot.

2) The coldest and hottest states may not always produce the maximum stress range. Think of a system where one leg is standby while the other is in operation. Switching between these legs might produce a greater stress range on the tee than the ambient to all hot case.

Hello everyone,

I’m analyzing a piping system under the B31.3 code, with the following conditions:

Saturated Steam
Operation (given):
P2 = 145 psig
T2 = 364 F

Design (calculated from: Tdesign = T Ope. + 50 F, as per my company design criteria):
T1 = 414 F
then, P1 = Psat@414 F = 273 psig

Minimal pipe Temp (installation) (given):
T3 = 41 F

Hydro Press (given) = HP = 217.5 psig
-----------------------------------------------
I had doubts about the appropriate load cases I should use, and I found this thread very useful.

I also read Richard’s “Satisfying Expansion Load Case Requirements”, on the COADE Newsletter - July 2003, and the B31.3 Appendix S, after reading Dave’s post at the end of the thread.

Based on what I’ve learned so far from those discussions, I’m proposing the following load cases:

L1 = WW+HP (HYD)
L2 = W+T1+P1 (OPE)
L3 = W+T2+P2 (OPE)
L4 = W+T3+P3 (OPE)
L5 = W+P1 (SUS)
L6 = W+P2 (SUS)
L7 = W+P3 (SUS)
L8 = L2-L5 (EXP)
L9 = L3-L6 (EXP)
L10 = L4-L7 (EXP)
L11 = L1-L7 (EXP) (“extreme” stress range)

However, I still have doubts, and I know I’m missing something important here, and perhaps have to do with the code. That’s what I come to you the more experienced people and experts, for help. My questions are:

1) I see that other fellows in the forum, like the friend Rajinder Singh, talk about T1, T2 and T3 but nobody considers the design pressure but only the higher operating pressure, in the stress calculations; not even the Appendix S considers Pdesign, Tdesign. However, the standard of my company states:
“Mechanical Design Temperature shall be used for:
Stress calculations;
Support load calculations;
Checking of pipe displacements at pipe hangers and supports to ensure continued support;
Sizing of spring hangers, based on multiple load case design, to accommodate highest possible travel.”

¿What should I consider, then? ¿What’s the right thing to do?

2) Can you give me your opinions, and insights on my proposed load cases, and/or come up with the right one?

3) Should L11 be = T1-T3 (EXP), as Rajinder proposed in the forum?

4) A very experiences colleague in a previous analysis of similar conditions to mine used the case: WNC+T3 (SUS), which seems sensate to me. What do you think about it? Should I use it as L7 instead of the one I proposed?

About the ASME B31.3:

"319.2.2 Displacement Stresses

(b) Overstrained Behavior. Stresses cannot be considered
proportional to displacement strains throughout a
piping system in which an excessive amount of strain
may occur in localized portions of the system (an unbalanced
system). Operation of an unbalanced system in
the creep range may aggravate the deleterious effects
due to creep strain accumulation in the most susceptible
regions of the system. Unbalance may result from one
or more of the following:
(...)
(4) variation of piping material or temperature in
a line. When differences in the elastic modulus within
a piping system will significantly affect the stress distribution,
the resulting displacement stresses shall be computed
based on the actual elastic moduli at the respective
operating temperatures for each segment in the system
and then multiplied by the ratio of the elastic modulus
at ambient temperature to the modulus used in the analysis
for each segment."

5) Should I understand “not the design temperature”?

"319.2.3 Displacement Stress Range
(…)
(b) While stresses resulting from displacement strains
diminish with time due to yielding or creep, the algebraic
difference between strains in the extreme displacement
condition and the original (as-installed) condition
(or any anticipated condition with a greater differential
effect) remains substantially constant during any one
cycle of operation. This difference in strains produces a
corresponding stress differential, the displacement stress
range, which is used as the criterion in the design of
piping for flexibility. See para. 302.3.5(d) for the allowable
stress range SA and para. 319.4.4(a) for the computed
stress range SE."

6) Shouldn’t it be the design temperature?

"(c) Average axial stresses (over the pipe cross section)
due to longitudinal forces caused by displacement
strains are not normally considered in the determination
of displacement stress range, since this stress is not significant
in typical piping layouts. In special cases, however,
consideration of average axial displacement stress
is necessary. Examples include buried lines containing
hot fluids, double wall pipes, and parallel lines with
different operating temperatures, connected together at
more than one point.

319.3.1 Thermal Expansion Data
(a) Values for Stress Range. Values of thermal displacements
to be used in determining total displacement
strains for computing the stress range shall be determined
from Appendix C as the algebraic difference
between the value at maximum metal temperature and
that at the minimum metal temperature for the thermal
cycle under analysis."

7) Max. Op. Temp or design Temp.?

"(b) Values for Reactions. Values of thermal displacements
to be used in determining total displacement
strains for computation of reactions on supports and
connected equipment shall be determined as the algebraic
difference between the value at maximum (or minimum)
temperature for the thermal cycle under analysis
and the value at the temperature expected during installation."

8) Max. Op. Temp or design Temp.?

Thanks so much for your valuable help.

Correction: L11=L2-L7. Excuse me.