open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi,
I tried a 2D case using FINITE VOLUME METHOD with the following details

TIME STEP = 0.1
VARIABLE TIME-STEP = NO
DESIRED COURANT NUMBER = 1

I found that the Time count for each iteration is varying which is not based on the given TIME STEP. I would like to know why does this happen?

Thanks
Sathya

Hello Sathya

You can not fix time step and Courant number in the same time. Courant number is used to compute time step which is necessarily variable.
the recommended set of keywords is the following:
TIME STEP = 0.1 (will be used only for 1st iteartion)
VARIABLE TIME-STEP = YES
DESIRED COURANT NUMBER = 0.9

For stability reasons, you might limit Courant number strictly less than 1.
with my best regards


Riadh ATA

Hi Riadh,
Thanks for providing the suggestion. I feel that the following detail explains one the problem which is being dealt is with FIXED TIME STEP. The DESIRED COURANT NUMBER =1 is default which comes from DICTIONARY file (irrespective of FIXED or VARIABLE TIME STEP).
TIME STEP = 0.1
VARIABLE TIME-STEP = NO
DESIRED COURANT NUMBER = 1

So my question is 'Though i had used FIXED TIME STEP with default DESIRED COURANT NUMBER, why does Time count for each iteration varies, which is not based on the given TIME STEP?

Sathya

Hi Sathya,

as Riadh said, you can not use a fixed time step and a fixed Courant number at the same time. Maybe the following equation explains it a bit more.


CFL = (u * dt)/dx


u is not constant, dx is constant.

Since the stability is limited by the transport of information between the cells, CFL should be set to 0.9 for the finite volume schemes.

Best regards
Leo