I'm confused as someone define cases just as follows.
these cases are for branch isolated from main pipe,
1 WW+HP HYD
2 W+D1+T1+P1 OPE design temp
3 W+D2+T2+P1 OPE ope temp
4 W+D3+T3+P1 OPE mini temp
5 W+P1 SUS
6 W+D2+T2+P1+WIN1 OPE
7 W+D2+T2+P1+WIN2 OPE
8 W+D2+T2+P1+U1 OPE
9 W+D2+T2+P1-U1 OPE
10 W+D2+T2+P1+U2 OPE
11 W+D2+T2+P1-U2 OPE
where D1，D2，D3 are displacements in main pipe of corresponding cases,(that is design,ope and mini temp), are these cases upward right?
should I add diplacement of sus case of main pipe to case 5?
just as following:
5 W+P1+D4 SUS (where D4 is displacement of sus case in main pipe)


another question: Is there any differece when the stress type of case 6 to 11 is OCC.
give me some advice, thanks!

Hi maingun
I think you have to add displacement for sustain case also, same as we consider it in settlement case.
For case 6 to case 11 occation case & we use it for restraint load check only so I feel there should not be any dfference between OPE & OCC. Mr. Richard can better clarify.
I will not like to make change against conventional practice.

Regards

Habib

Whoa there! If you are separating branches for ease of analysis, you need to transfer movements from corresponding load cases from the main run to the separated branch. Usually, the sustained load case displacements at the branch connection are small enough we can disregard them, but they should be checked.

... and you'll miss the interaction between the two systems. You need to be "very careful" breaking systems like this.

Thanks for your reply!

another 3 qustions:
1) for case 1
1 WW+HP HYD
should I add displacement to this case like case 5 in branch?

2) should these displacements of main run we considered just are normal
operating case or should be corresponded to every case,
I list these cases for clear,
for main run displacement in intersection node
1 WW+HP HYD D11
2 W+T1+P1 OPE design temp D1
3 W+T2+P1 OPE ope temp D2
4 W+T3+P1 OPE mini temp D3
5 W+P1 SUS D4
6 W+T2+P1+WIN1 OPE D5
7 W+T2+P1+WIN2 OPE D6
8 W+T2+P1+U1 OPE D7
9 W+T2+P1-U1 OPE D8
10 W+T2+P1+U2 OPE D9
11 W+T2+P1-U2 OPE D10

Which cases for branch is right, a) or b)
a)
1 WW+HP+D11 HYD D11
2 W+T1+D1+P1 OPE design temp D1
3 W+T2+D2+P1 OPE ope temp D2
4 W+T3+D3+P1 OPE mini temp D3
5 W+P1+D4 SUS D4
6 W+T2+P1+D5+WIN1 OPE D5
7 W+T2+P1+D6+WIN2 OPE D6
8 W+T2+P1+D7+U1 OPE D7
9 W+T2+P1+D8-U1 OPE D8
10 W+T2+P1+D9+U2 OPE D9
11 W+T2+P1+D10-U2 OPE D10
b)
1 WW+HP HYD
2 W+T1+D1+P1 OPE design temp
3 W+T2+D2+P1 OPE ope temp
4 W+T3+D3+P1 OPE mini temp
5 W+P1+D4 SUS
6 W+T2+P1+WIN1 OPE
7 W+T2+P1+WIN2 OPE
8 W+T2+P1+U1 OPE
9 W+T2+P1-U1 OPE
10 W+T2+P1+U2 OPE
11 W+T2+P1-U2 OPE
3)As shr said, stess type of cases 6 to 11 defined as OCC rather than OPE is conventional practice, am I right?

Hi maingun

1) Wind & earth quake is an occasional load.
2) Stress for wind & earth quake to be considered as occasional stress.
3) Load case you mention above is used for restraint load checking not for stress checking -B31.3 ( expect appendix P 31.3) hence stress type mention as OCC or OPE will not make any difference as far as load is concern.
4) Load case that is used for stress checking should be specify clearly as SUS, OCC or EXP since allowable stress criteria is different for different cases.
5) Displacement should be enter all those cases you mention though displacement for sustain case & hydro case should be very minimum.
6) As Mr. Richard mention be very careful to break system like this. Because header displacement affect branch & same time branch also affect header so clear picture can not be drawn unless combined analysis is done. When branch size is very small compare to header system can be break like this considering concept “ Dog can shake tail , tail can not shake dog”


Regards
Habib

dear shr,
thanks for your quick reply, still i am not clear with following cases,
OCC LOAD for branch(red colour),
Which cases for branch is right, a) or b), according to your relpy, a) should be choosed, but still I think b) is also right because in this case ,occ load have considerded the ope load.
please shed some light on me, thanks!

a)
1 WW+HP+D11 HYD D11
2 W+T1+D1+P1 OPE design temp D1
3 W+T2+D2+P1 OPE ope temp D2
4 W+T3+D3+P1 OPE mini temp D3
5 W+P1+D4 SUS D4
6 W+T2+P1+D5+WIN1 OPE D5
7 W+T2+P1+D6+WIN2 OPE D6
8 W+T2+P1+D7+U1 OPE D7
9 W+T2+P1+D8-U1 OPE D8
10 W+T2+P1+D9+U2 OPE D9
11 W+T2+P1+D10-U2 OPE D10
b)
1 WW+HP+D11 HYD D11
2 W+T1+D1+P1 OPE design temp D1
3 W+T2+D2+P1 OPE ope temp D2
4 W+T3+D3+P1 OPE mini temp D3
5 W+P1+D4 SUS D4
6 W+T2+P1+D2+WIN1
7 W+T2+P1+D2+WIN2
8 W+T2+P1+D2+U1
9 W+T2+P1+D2-U1
10 W+T2+P1+D2+U2
11 W+T2+P1+D2-U2

Hi maingun

Forget case OCC or OPE.

Think what could be the situation for combined analysis try to represent the same as possible for break system. So case a is you mention should be more appropriate.

Regards

Habib

Dear shr,
Now i am clear, thanks very much!

Another question coming, about the cases I considered above, how can I define node displacements with more than 9 cases, In CAESAR II I can only define 9 displacements at most. Is there any way to deal with this situation(cases more than 9)?
thanks for your reply!

You can have 9 displacement vectors - that's it.

Hi maingun

Simple make two caesar file, one with wind another with seismic.
If possible better make combined system.

Hi shr,
thanks for you help!

Hi everybody,
I was viewing this interesting topic which has been my concern for a long time. As I see you are not worried about transfering the moments between the two systems while braking them apart. Would you justify this please.

In my openion forces,moments and movements shall be transfered from one system to another of the two broken up systems.

Dear Habib what is your openion?

Regards,
Aslam

Hi Omega/Aslam

Ideally (if possible to find out) we should transfer force, moment, movement & stiffness of that point in both the system.
It makes both the system as if combined system.

Here force & moment means interactive force of the system, when system is in join position. Both system produce effect on each other.
Generally it is not a easy task. That's why we can not break system arbitrarily.

We break system when header size is very big, branch size is quite small.
We assume branch is connected or not connected will produce same displacement in header calculation. In that case only displacement from header calculation, we can import in branch calculation. we can ignore force moment & stiffness.

If system is highly critical one, better to go for combined analysis.

Regards

Habib

Dear Habib,

Forces,moments and movements could be found from header analysis report to include them into branch analysis but how will you get stiffness values from header analysis report and most important how you will input it in caesar II.

The header-branch analysis as two seperate analysis is a common situation encountered by every stress engineer.

The general practice is first to analyse header portion considering branch disconnected with it and note down the displacements at branch connection point.While anlysing branch portion at that time define the earlier noted displacements at the branch connection with applicable SIF and evaluate stress at the connection point.

Stiffness is a system property and depends on modulus of elasticity,cross-sectional area,thickness and length of piping element for any simple piping system including header-branch configuration.

We incorporate stiffness of the system by modelling branch piping with portion of the header upto first support on either side of branch.

My question is why we do not consider forces and moments together with displacements to evaluate stress at the connection point.

Explanation supported theoritically will be highly appreciated.

Regards,
Aslam

Hi Aslam

Here force & moment should be local branch element force & moment.
We have to model small element for branch & from local element force of that branch we need to take the force & moment data to input in branch calculation. We should not pick data from header side.
I understand that value will be very minimum.
Stiffness we ignore will produce more conservative result hence no issue. Displacement data make the system as displace anchor point.
Yes sometime we model upto two three additional support that produce more realistic result.

You mention correctly SIF we must feed for branch calculation.

I summarized like
1)Branch element force & moment very small ( taken from header calculation)
2)For branch calculation we are using displace anchor hence conservative.

So we can forget force & moment , only displacement & SIF data require to go for branch calculation.


Regards

Habib

Hi Aslam,

As I understand, We only need to add displacements on tie-in / break points. Displacement already represent acting forces and moment on this point. By adding displacement & F&M, it will be concerned twice.

Regards,
TGM

Either displacements AND rotations, OR, forces AND moments.

But remember that your system will also produce "balancing" loads. These loads from your system may change the position (and loads) of the system represented by that position of load set.