Steady calculation with CS

Hello all,

I'm here again (because you're very helpful !). I want to run a calculation for a steady flow. So far it doesn't seem difficult, it's just some things aren't really clear in my head.
Basically, a steady flow is calculated on several iterations (numerical iterations which don't have any physical meaning). For each iteration, the solver manage to obtain converged variables on the basis of the convergence criteria given in the equation parameter tab. Then for a steady flow we want to have converged variables between two iterations (ie the difference between two successive iterations for each variable must be lower than a certain value).
In C_S, we can only give a number of iterations, and a convergence criteria to reach to start to the next iteration. Why can't I give the same kind of criteria to stop the calculation when the steady state is reached ? When do I know that the steady state is reached ? Can I see the evolution of the difference of the variables between 2 iterations and thus see when the steady state is reached ?

I would be very grateful if someone could explain to me how C_S manage the steady flow calculation...

Alicia wrote:

... When do I know that the steady state is reached ? Can I see the evolution of the difference of the variables between 2 iterations and thus see when the steady state is reached ? ...

Interesting question, actually. I do not think that - at least for turbulent flow - the term steady state applies at all physically, but it may apply statistically - i.e. considering average values. My thoughts on the topic at hand is to monitor how the condition changes in iteration history using the "MONITORING POINTS MANAGEMENT" setup and then to fiddle with the relaxation coefficient. But no real experience ...

I would also really like to see an explanation of convergence criteria with Code Saturne. I seem to recall looking at the number of iterations after a simulation had completed (in some text file) and there were more iterations than I had requested (I am quite unsure about this). It just seemed to me that Code Saturne does not just run the number of iterations requested, it has some algorithm that uses this specification.

I asked a sort of related question on the Code Saturne forums (code-saturne.info/products/code-saturne/...eral-usage/122810249). This is the response I received:

In the "listing" file created in the RESU directory, you get some information about the convergence of the code. At each time step (or at the NTLIS periodicity you specified), the code will write a block "INFORMATION ON CONVERGENCE" with info on each variables, especially:

* N_iter: nb of iterations used to solve the linear system

* Norm. residual: normed residual after solving the linear system (should be lower than EPSILO=1.e-8 which is the criterium for stopping the iterations)

* derive: roughly corresponds to |u(n+1)-u(n)|/|u|

If you increase the verbosity of the code for certain variables, by putting IWARNI(IVAR)=2, 3 or 4 in usini1.f90, you will have much more information, in particular the residuals at each iterations of Jacobi or conjugate gradient.<br /><br />Post edited by: Matthew Bondy, at: 2010/03/29 03:11