I had never given thought to the Total Stress Range needed for pump systems where one pump is operating and one pump is standby until a recent posting that I can't find now. I have set up the following load cases for my system. I don't know which would be the correct Total Stress Range - Case 10 where the Expansion Cases are added or Case 11 where the Operating Cases are subtracted. The results of the two cases are significantly different.

CASE 1 (HGR) W
WEIGHT FOR HANGER LOADS

CASE 2 (HGR) W+T1+P1
OPERATING FOR HANGER TRAVEL

CASE 3 (OPE) W+T1+P1+H
OPER T1 All @ 800F

CASE 4 (OPE) W+T2+P1+H
OPER T2 P-3B @ 800F, P-3A @ 70F

CASE 5 (OPE) W+T3+P1+H
OPER T3 P-3B @ 70F, P-3A @ 800F

CASE 6 (SUS) W+P1+H
SUSTAINED

CASE 7 (EXP) L7=L3-L6
EXP T1 All @ 800F

CASE 8 (EXP) L8=L4-L6
EXP T2 P-3B @ 800F, P-3A @ 70F

CASE 9 (EXP) L9=L5-L6
EXP T3 P-3B @ 70F, P-3A @ 800F

CASE 10 (EXP) L10=L8+L9
EXP T2+T3 Total Stress Range
Keep/Discard: Keep
Display: Stress
Combination Method: SCALAR
Flg Analysis Temp: None

CASE 11 (EXP) L11=L4-L5
EXP L4-L5 Total Stress Range
Keep/Discard: Keep
Display: Stress
Combination Method: SCALAR
Flg Analysis Temp: None

With L3, L4, L5 & L6 each being a "state" of your piping system, and assuming nonlinear restraints, I would consider the following expansion stress ranges:
L3-L6, L4-L6, L5-L6 then
L3-L4, L3-L5 and
L4-L5

Now, I did not say all these cases required analyiss, I just said they should be considered. Several may be insignificant.

I would use ALGEBRAIC as the combination method.

I wonder how many engineers consider the combined effect of all these independent range calculations? The Code (B31.3) includes a paragraph on accummulated damage from different sets of (fatigue) cycles. If you have more than one expansion stress range that's high with respect to its allowable stress range, then this may be important.

Not inferring that Mr. Diehl provided an inadequate answer here is a link to additional information that I have found helpful:

http://coade.com/newsletters/jul03.pdf

Sir Dave,

Its been a while since last discussion with Sir Richard Ay in this subject and yet I got this big ? still hanging on my head..

I would like to raise a question with Algebraic combination of L3-L4. My question is all about the strain of the header with same temperature. I still don’t know, how caesar treat this matter.. will caesar retain the thermal strain of the header or it cancel out.. Can you please give an overview on how caesar do this?

Sorry if this sound stupid but it really convince me of not using this "critical" evaluation..

Regards!

That L3-L4 case will examine the change (or range) between those two states - the state of the piping system under L3 conditions and the state of the piping system under L4 conditions. This is just one of (apparently) several cycling events (others are L4-L5, L3-L6, etc.). Each range will reduce the remaining life of the piping system by accumulating this fatigue damage. The piping code offers a simple accounting of these various ranges by converting lower levels of stress change into equivalent cycles at the highest calculated stress range. In this manner, the highest calculated range will be the only stress range examined but used with the adjusted number of equivalent cycles, N. Note that the largest stress range may change from node to node - L3-L4 might cause the largest expansion stress at node 30 while L5-L6 might produce the largest at node 130.

I do not believe most engineers do this accummulation. They simply check individual expansion stresses.

So, I have no issue if L3-L4 shows a stress change of zero in the leg that is hot in both L3 and L4. That cycle pattern will not add to the equivalent cycles in that leg.

I doubt there are very many pumps systems that experience thermal fatigue damage while in compliance with the allowable nozzle loads.

Getting the number of cycles defined for what is not normally considered a cyclic system would be a challenge.

Good point. In satisfying pump limits you are probably guaranteed a low expansion stress range in the vacinity of the pump.

I guess I was trying to hijack this thread to talk about evaluating multiple thermal ranges.

Sir Dave thanks for the reply.

It is not my intention to question the example #3 of App. S of the code but how it is treated/executed in CaesarII which at all case one common leg appear with same temperature. Does CaesarII minus both thermal strain? If not.. How is it solved?

NozzleTwister thank you for additional headache

It will appreciated if you will share the discussion of stress range factor increase for this subject...

Regards!

For structural response (displacements & rotations, forces & moments) CAESAR II performs the combination as listed. For example, for L5:L3-L4, the X displacement for load case 5 is equal to the X displacement from load case 3 minus the X displacement from load case 4.
Because of Code requirements, stresses may be handled differently, that's why we offer "algebraic" and "scalar".

Sir Dave your reply is really appreciated.. I would like to follow-up my question BUT I will save this one later when I’m prepared..

I will be happy if you can send me an algorithm in evaluation of the said topic, if it is possible.

THANK YOU VERY MUCH...

Regards,