B31.3 Allowable stress for stress range.

Hi all,
I have a case where one branch is at high temperature while adjacent pipework cycles from cold to hot.
When calculating stress range, the allowable stress range at branch is calculated using SA=f(1.25Sh+0.25Sh) rather than f(1.25Sc+0.25Sh). Is it due to the fact that branch is at high temperature in both cases?
The Sc is defined as "basic allowable stress at minimum temperature expected during the displacement under analysis"
There is difference in opinion at our place regarding this. In my opinion it is correct to use SA=f(1.25Sh+0.25Sh)but some say it should be f(1.25Sc+0.25Sh) Sc being at ambient temperature.
I would like to have your expert opinion on this.
I have attached the sample CAESAR model herewith.

Many thanks and regards

MW

Any comments?????

I could not open the file due to versioning issues.

However, I would have to argue that the connection should be analyzed at the hottest component of the juncture that's assigned without any proper justification for using a lower temperature.

In addition, I would caution that if the temperature differential is great (say 100s of degrees difference), you have significant internal stresses within this component that won't be addressed by CAESAR.

Depending on the properties of the fluid and the relative flows, you may also have thermal bowing in piping downstream of the connection.

As a follow-up, I ran a test case. I have three sets of 2 tees.

2 tees are at 100°. (nodes 10-20, 20-30, 20-40) & (110-120, 120-130, 120-140)
2 tees are at 100° run and 800° branch. (add 200 to previous node numbers)
2 tees are at 800°. (add 200 again)

All 6 tees are anchored on one run side.

3 tees are restrained rx, ry, and rz on the branch end, with a force/moment applied to the opposite end of the run as F1= 1000x, F2=1000y, F3=1000z. M4=1000x, M5=1000y, M6=1000z. (10-40, 210-240, 410-440.)

3 tees are restrained and a load applied as the reciprocal as described in the previous paragraph. (110-140, 310-340, 510-540).

Because nodes X20 (020, 120, 220 etc) show up three times each in element definition, stresses are calculated three times for each X20 node, each using the temperature of its parent element.

So, yes, allowable stress is calculated at cold, and yes, allowable stress is calculated at hot. Stresses are then calculated for each end individually, and the maximum will apply for that node.

Michael,
Thanks for your reply.

I have attached the screen shots of T2 and T3 cases.

For T2 - Sustained case: Allowable stress is 27650 psi

For T3 - sustained case: Allowable stress is 27650 psi

For T3-T2 (range): Allowable stress is 15902 psi.

For T2 Allowable = f(1.25Sc + 0.25Sh (at 520C))
For T2 Allowable = f(1.25Sc + 0.25Sh (at 520C))
= f(1.25*20000 + 0.25*10601)=27650

For T3-T2 Allowable = f(1.25Sh (at 520c) + 0.25Sh (at 520C))
= f(1.25*10600+ 0.25*10600)=15902
Sc = 20000psi
Sh(at 520C0) = 10601 psi

What may be the reason for lower allowable for the range.

Kind regards

MW

This is possibly a new one on me.

I assume node 480 is the tee in the vertical. If so, this makes sense if Sh@204°C is also 10600 psi, but that sounds a bit off.

If node 480 is the branch connection on the left, it remains 510° between these two temperature profiles, and I assume L8 is an ambient sustained case, thus S=20000 would be applicable.

Michael,

The node 480 is on the west side (north being the -X). Yes the tee is at 520C in both temperature cases.
Regards

MW

The difference is due to the value used for Sc. By definition Sc is the allowable at the coldest temperature of the range. For the first two load cases (T2-ambient) and (T3-ambient), Sc is 20000. For the final range (T2-T3), Sc is 10601.