CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   using steady state results for unsteady flow (http://www.cfd-online.com/Forums/main/10823-using-steady-state-results-unsteady-flow.html)

siva appanna February 18, 2006 06:28

using steady state results for unsteady flow
 
i edi performed a incompressible steady state flow for a centrifugal compressor...how am i suppose to use these results for compressible unsteady flow simulation using segregated solver?

Ravi Kiran February 18, 2006 09:31

Re: using steady state results for unsteady flow
 
Hi Siva

U continue the simulation from where u have stopped by changing the flow properties of the fluid to compressible ones and the inlet and outlet conditions as accordingly

this would mean that the earlier flow developed by incompressible analysis would be initial value of the domain for next iterations

siva appanna February 18, 2006 21:54

Re: using steady state results for unsteady flow
 
hi ravi kiran!

thanx for help ..my simulation at 20k rpm manage to converge finally...now . i'm facin with another problem...i ran my simulation at 70 k prm using segregated solver for imcompressible flow...after 5000 iterations , my simulation din converge...my under relaxation factors were at their default values....can u advise u me on how i can solve this problem?what are the issues that are relate to convergence of a simulation? thanx in advance for ur help


Ravi Kiran February 20, 2006 06:29

Re: using steady state results for unsteady flow
 
Have U try with coupled solver for your problem

and regarding the relaxation parameters u have to judge by experimentation with smaller mesh

in case of coupled solver also u can continue with previously simulated data

and while using coupled solver u can control the solution with smaller Courant number initially and later increasing the courant upto 10 or even 100

what it does is the it first makes the mass balance and later goes for convergence of the flow field fluxes(i have found in some cases which i did)

ok try this one if u have time

all the best bye Ravi


All times are GMT -4. The time now is 15:23.