CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   Timescale (http://www.cfd-online.com/Forums/cfx/23457-timescale.html)

Luke December 21, 2006 23:26

Timescale
 
Can someone please explain to me the use of a timescale in CFX for steady state problems..... Is'nt a timestep needed only to solve transient problems? What is the significance of this timestep in steady state problems....?

Thanks in advance

Phil December 22, 2006 07:16

Re: Timescale
 
The physical timestep is a means of underrelaxing the solution in a physically meaningful way as it iterates toward the steady solution.

Omer December 24, 2006 08:48

Re: Timescale
 
In simple words. When we give inlet conditions and outlet conditions, they are only at the boundaries. so the rest of the domain is empty so to speak at time zero. So the solution has to progress in time no matter what, even if it has to go to steady state solution. The Steady state itself , means the variations with time becomes zero (almost), but it is not neccesarily at a physical time zero.


aero December 26, 2006 07:35

Re: Timescale
 
Hi,

i think, this concept is covered in tutorials.

Time scale can be given by

distance travel/inlet velocity.

check the tutorial of Static Mixer for more details.

bye aero

Bak_Flow December 30, 2006 13:49

Re: Timescale
 
Hi Luke,

there is a nice non-commercial CFD explaination of this in Versteeg and Malalasekera's book an Introduction to Computational Fluid Dynamics, 8.8

They show nicely how under-relaxation and pseudo-transient advance to steady state have a very close relation. They give the formula. ... either believe me or refer to the book.

"This formula shows that it is possible to achieve the effects of under-relaxed iterative steady state calculations from a given initial field by means of a pseudo-transient computation starting from the same initial field by taking a step size that is given by eq 8.48....The pseudo-transient approach is useful for situatyions in which goveringinewuqations gives rise to stability problems eq buoyand flows, highly swirling flows and compressible flows with shocks."

It has been my experience as well that what I will call "hard to solve problems" converge more reliably and faster with this method. Unfortuntatly, some problems have a wide range of time scales and are not well suited to this method as there is some stability limit governed by the smallest time scale in the problem which prevents using a large time step.

Best Regards,

Bak_Flow


All times are GMT -4. The time now is 02:07.