open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi,

I am trying to run a 3D model, steady-state case, for the fluxes through a reservoir. I have been trying some configurations but without success in getting stable results or in some cases no results at all, as the solver blows without a solution.

I kindly ask for any suggestions. I am attaching the steering and boundary conditions file. I a nutshell, I am using the LIPS scheme, tidal flats, k-epsilon model and keywords such as preconditioning, accuracy for solver and others are set to default.

This is a screenshot of the results that I am having, naturally, my first thought would be to increase the maximum number of iterations, but I am not completely sure if this could work as I do not get any convergence with the default values.

Also, I have a question regarding tidal flats. In the steering file I have the keywords: OPTION FOR TREATMENT OF TIDAL FLATS = YES and TREATMENT OF NEGATIVE DEPTHS = 2. On the other hand, there is also the keyword MINIMAL VALUE FOR DEPTH, but it is my understanding that this works when TREATMENT OF NEGATIVE DEPTHS = 1. I just want to confirm that.

Many thanks in advance.

Karla

Hello Karla,

Indeed, when you have a warning message like yours
"GRACJG (BIEF) : EXCEEDING MAXIMUM ITERATIONS:... RELATIVE PRECISION:..."
you can increase key keyword "MAXIMUM NUMBER OF ITERATIONS FOR..." associated to the step to solve. You also can relax the keyword "ACCURACY FOR..." associated to the step to solve if too hard to converge.
Perhaps you need to decrease the time step.

If you try to model a reservoir, you can also try the set-up of the $HOMETEL/examples/waqtel/heat_exchange example.

The keyword MINIMAL VALUE FOR DEPTH is used with option OPTION FOR THE TREATMENT OF TIDAL FLATS = 2 (which is not recommended as not conservative and only may work in sequential run) or when masking elements.

Hope this helps,

Chi-Tuan