CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   how to increase time step size? (http://www.cfd-online.com/Forums/fluent/33715-how-increase-time-step-size.html)

co2 May 12, 2004 13:28

how to increase time step size?
 
hi all,

I am solving a natural convection unsteady problem. I am getting good convergence in 20 iterations per time step for a step size of 2 sec. --- so I decide to move to 3 sec time step size -- That results in a problem -- continuity residual does not seem to go down. it hovers around 1.6e-4 yes, i have nice stringent convergence criteria for an accurate solution. Can you tell me how i can run my simulation at 3 sec time step size? what urfs do i need to bring down ? both my presure and vel urfs are at 0.1 already.

thanks, co2

Tom May 13, 2004 03:55

Re: how to increase time step size?
 
Are you using the PISO scheme for pv coupling? If so you can set your urf much higher. How good is your mesh?

ap May 13, 2004 04:03

Re: how to increase time step size?
 
I'd try to increase the underrelaxation factor for pressure, usually this leads to a faster convergence.

If residuals stay high, keep the time step you're using.

Hi :)

ap

co2 May 13, 2004 12:08

Re: how to increase time step size?
 
my mesh is good with about 27000 cells for a 2D axi problem. I am seeing good results.

I am using PISO for pv coupling -- but trying to increase p and v urfs does not seem to help. Thus I guess there is no possibility of increasing the time step size

Venu gopal May 14, 2004 06:01

how to increase time step size?
 
Haai, This is venugopal from IIT MADRAS. Regarding to your problem, for natural convection always you have to go for less time step size, in the sense below one. If you are increasing the time step above one or more, the discretization error increases, and some times you will face the problem to get converged solution as what you told. So, you go for less convergence, if you are going below 1 sec, you will get good results. Increasing time step means, increasing erros. Bye Venu Gopal

co2 May 14, 2004 11:00

Re: how to increase time step size?
 
thanks everyone for your reply. I think my mesh is good. The only skewness is because of the meshing of solid zones in my model -- I have to model the thin steel walls of my tank as I have to put heat source terms in them to account for external sun radiation.

91 % cells have equiangle skewness less than 0.1

kethireddy May 17, 2004 07:25

Re: how to increase time step size?
 
Hi This is kethireddy from RWTH-Aachen.Germany. U reduce the concergence criteria for the equation u solving.If u want u can reduce the no of time steps in one second.(in the iterate menu for unsteady state calculations) ........ kethireddy


All times are GMT -4. The time now is 20:14.