Hello again Everyone,
Can anybody give me an enlightment about when do we have to do Non-Liniear analysis in CAESAR II, I understand about single direction restraint, gap and friction if any of this in our modelling we have to do Non-liniear analysis, but i have seen my senior doing Liniear analysis even if he have a gap, friction and single direction restraint in his modelling, so i need another explaination when do we exactly have to do a Non-Liniear analysis in CAESAR II, coz if the system is so complicated contained of line with 2 piping spec., 3 pumps and PSV with gap and single directional restraint in the system and have a seismic activity and wind as an example maybe it will be so many case that i have to set in load case editor,it could be more than 99 load case.

If you have non-linear boundary conditions, CAESAR II will perform a non-linear analysis (and you can't stop it). The only way to avoid a non-linear analysis (and the inherent interations) is to manually linearize the model. (This isn't the best course of action though.) Manually linearizing the model involves removing the non-linear restraints (which could change the outcome of the analysis.)

Dear sir,
If Caesar II will automatically do the Non-liniear analysis, why in CAESAR we have to build a different load case for non liniear boundary conditions? i mean as an example for seismic in non liniear we have to do like this (Just as an example):
L2: W+T1+P1
L3: W+T2+P2
L4: W+T1+P1+U1
L5: W+T1+P1+U2
L8: L2-L4
L9: L2-L5
because i have see in AUTOPIPE we don't have to build a case for Non-liniear analysis, AUTOPIPE will do it for you, only there is a User Defined Combination if you want build your own case.

1) So i want an enlightment when we have to build Non-liniear case in CAESAR II?
2) If in my system modelling i have a gap or friction or single directional restrait like +Y, should i build the case like an example above or just drag and drop the acceleration (U) to the load case editor?

Why do we suggest this approach to determining the seismic component of system response with nonlinear restraints?
Using your load case examples. L2=operating state, L4=Operating plus seismic #1 (U1) & L8=L4-L2 (I reversed your loads here).
Let's say you have a +Y support that carries a downward load of 1500 lbf in L2 and a downward load of only 500 lbf in L4. The pipe never disengages from the support. L8 would show this. However, if you would run U1 by itself, that 1000 lbf upward load would probably result in liftoff at that support and the moments and streses would be different, maybe quite so.
Evaluating seismic (U1) alone would not reflect actual system response.
I'll make it a little more complicated: if U1 is X and U2 is Z, you should not only start with operating+U1 but also check operating-U1.

So if i get your point here Mr.Dave, it gives me a conclusion like this:

IF I have +Y Restraint in my sistems in particular position, so does the gap and friction, first i'll run the analysis with Liniear analysis, if in Seismic case or OPE case i found that any of this support lift off it means that i have to build the case in Non Liniear analysis like COADE suggested. if none of the support lift off, so i can just do yhe Liniear analysis.

Correct me if i'm wrong.