CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   STAR-CCM+ (http://www.cfd-online.com/Forums/star-ccm/)
-   -   Problem with Convergence (http://www.cfd-online.com/Forums/star-ccm/110327-problem-convergence.html)

badger1 December 8, 2012 08:10

Problem with Convergence
 
2 Attachment(s)
I am having a problem with my Starccm simulation in that it will not converge to a final solution. the residuals are very high are the solution never seems to reach a steady state where everything is constant it just keeps moving.

I am simulating a turbine blade and I am using the following models:
gamma ReTheta
K-omega
Steady
Turbulent
2D
Coupled
All Y+ treatment
Ideal Gas

I am hoping that there is something I can do in order to make it converge and reach a solution. I have attached some pictures so that you can tell what I mean. The solution is periodic which explains a little better the scaler plot.

thanks

Attachment 17510

Attachment 17511

luizhferrari December 11, 2012 19:08

Oi, tente diminuir o fator de sobre-relaxamento, o default é 0.9, tente abaixar para 0.5, isto poderá deixar problemas de tempo em sua simulação, caso não, pode ser problema de modelo mesmo, tente colocar uma sonda e monitorar por isto.

Quote:

Originally Posted by badger1 (Post 396409)
I am having a problem with my Starccm simulation in that it will not converge to a final solution. the residuals are very high are the solution never seems to reach a steady state where everything is constant it just keeps moving.

I am simulating a turbine blade and I am using the following models:
gamma ReTheta
K-omega
Steady
Turbulent
2D
Coupled
All Y+ treatment
Ideal Gas

I am hoping that there is something I can do in order to make it converge and reach a solution. I have attached some pictures so that you can tell what I mean. The solution is periodic which explains a little better the scaler plot.

thanks

Attachment 17510

Attachment 17511


siara817 December 13, 2012 04:09

Seems to me that the mesh is coarse. But before that try to change physics models from second order to first order. See if it gets converged or not. if yes, then change to second order after for example 2000 iterations.

badger1 December 18, 2012 05:54

I tried to change both the order from 2nd to 1st and also the couant number and also the relaxation.

The simulation converges fine if you use segregated and constant density, but if you use coupled and ideal gas it does not and gives stupid results and never reaches a steady state solution.

I need to use coupled and ideal gas to match an experiment that someone else has done.

siara817 December 20, 2012 14:23

I think if you want to try coupled, you need to reduce more the relaxatoin factor. It will take certaily longer to converge, you need to be patient.


All times are GMT -4. The time now is 18:20.