Dear all:

I was trying to model a pretty huge pipe with lots of bends and all and with an inlet velocity of 27 m/s. the total span of the pipe on x dir would be around 10m and y would be around the same. i was trying to model a part of a pipe in a power plant. the problem was that i was getting divergence when i used an unsteady implicit solver..any thoughts...or is it just that the whole system is just very unstable one??

thanks in advance

this will need very good mesh i guess. i will say system is quite wild!

hi derick i think u dont have to model the whole pipe at same time. try model a type of geometry (means pipe of certain dia with certain bend angle)one by one. and for similar type just add the losses etc.

i hope i made my point clear.(divide and rule)

Try one or all of these to see if it works - smaller time-step size, smaller under-relaxation factors (segregated solver) or CFL number (coupled solver).

thanks guys for ur help..i will surely try all ur suggestions. hope they work...thanks once again guys