hi pipers,
i hav seen a situation that header is connected to 5 branches and the branches consists of bursting disk at their ends .so my question is how to analyse the system ,i hav to analyse the system into 5 parts+1 part has header or else i have to analyse the system as whole.

Depends how big system it is. If really big one you can do part by part. One more point is what is the header & branch size. Say header is 24 " & branches 4". System can be break. At all branched interface point feed displacement data from header stress calculation report along with SIF at that point.
If there is no much difference between header size & branch size then you can not break the system at header branch interface, you have to find some other suitable point at branches to put anchor & isolate systems.

I have a sneaking suspicion that you are looking at your system hydraulically. CAESAR II is a structural analyis program and not a hydraulics package. If load or deflection can pass through a "boundary" it is not a structural boundary and its effects must be considered. A burst disc or a closed valve is not necessarily a boundary or end of your CAESAR II model.

Manjula,

If you do a pipe stress analysis, it is always better to do the whole system analysis in one file. I don't like it either because it is a pain waiting for the whole model to regenerate everytime you change something.

You should not rely on your piping isometrics are basis on where to isolate stress analysis, always have your P&ID beside you.

My concept is that I only make another stress run for branches having a 1:25 ratio with section modulus of the main header line. Because, the effects of displacements/movements of the smaller line won't do much contribution to the main header line. But, keep in mind that you need to input all necessary data for the main-branch connections such as SIFs, allowables, etc...

Goodluck!

this sounds like a clean Pharma relief header.

the burst disks at the end of the branches close to the header are there to prevent backflow in case of down stream (header ) relief and they are typically ignored in an analysis because they are below 1barg.

i could be wrong