Load cases combination

Hi to all!!

I've a big problem and I hope someone will help me.
I use caesarII v.5.10 build 080512

The problem is: I'm trying to do a complex simulation in which I have to consider displacement of structures, inertial accelerations.
The result I obtain, for my opinion, is affected by errors.
Below you can see load cases combination and result:

FY (force in axial direction)
L1=W+P1 64N
L2=W+P1+T1 1648N
L3=W+P1+D1 -11295N
L4=W+P1+T1+D1 9857N
L5=D1 -4852N
L6=T1 593N
L7=L2-L1 1712N
L8=L3-L1 -11231N
L9=L4-L2 -11505N
L10=L4-L3 1439N

W=weight
P1=pressure
T1=temperature
D1=displacement

Note combinations shown above are worked out to obtain T1 and D1.
I' think of L6,L7,L10(to obtain T1) and L5,L8,L9(to obtain D1) should have respectively same results, but it'isnt.

Why?

All results relate to the same node which is an ANC.

PS: I've also seen axial force on valves is much more high then in a node in which there is a restrain (i.e. valve in Y(axial)direction and the restrain in Z(vertical) direction).

If your system includes non-linear boundary conditions (directional restraints, gaps, or friction), cases 6 and 7 will not yield the same results. This is exactly why we suggest case 7 is the better methodology.

Thanks for your reply mr. Ay.

My system includes non-linear boundary conditions (directional restraints, gaps, or friction); case L5 or L6 don't yield good results?

If it's so, I will not consider these cases, but I don't understand the cause results are fake.

Anyway I'll take in account cases L7(for T1) and L8 (for D1) because you suggest they are the best metodologies, but why cases L9 and L10 which are similar to L8 and L7 respectively haven't the same results?

The reason is because L5 and L6 take the difference from ambient to the load condition, while L7 and L8 take the difference between the SUS loading condition and the OPE loading condition, which can be quite different because of how the pipe has moved in the two conditions when nonlinear restraints (or friction) are involved. For example, say the SUS load case closes a gap to one side and the OPE condition has closed the gap to the opposite side. In this case the difference between them will give a displacement of twice the gap. Your T1 and D1 load cases would only see half of this displacement, resulting in an erroneous stress condition. We want to see the effects of strains (thermal and displacements) in the presence of other loads in the system, not completely isolated from the realistic loads.

Dear Italtecnica,

Just want to add links to previous posts:

http://www.coade.com/ubbthreads/ubbthrea...=true#Post16253
http://www.coade.com/ubbthreads/ubbthrea...=true#Post14855

Just for anticipation, I also want to paste Sir Loren Brown's point from second link.

-----------------------------------------------------------------------------------
"ONLY stress as the Output Type for your combination load cases (this include EXP as well). Displacements, Forces and Moments are NOT REAL in combination load cases"
-----------------------------------------------------------------------------------

L7 to L10 are kinds of combination load cases. L1 to L6 are the native load cases.

Dear Richard,first of all, happy new year, I understood thet case L7 is suggested. In this case I appraise the pure thermal condition. I don't understand,however, why haven't I to take in account L10? L10 is the best for me because it comes from the complete operating condition (L4).
Please remember that D1 represents structure displacements. If it were due to thermal displacement (e.g. nozzle) what does it change?
Regards

I agree, you should look at case 10. All I said was that case 6 isn't as correct as case 7.

For the cases listed above, I would eliminate cases 5 and 6. All the others are ok.