open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi,

I have a quick question about the "exceeding maximum iterations" error message. My simulation runs smoothly with no indication of a problem, until it quickly destabilizes in 2 time steps and stops running.

I have looked through other threads and adopted all the parameters Jean-Michel recommended:
TREATMENT OF THE LINEAR SYSTEM : 2
SOLVER : 1
FREE SURFACE GRADIENT COMPATIBILITY : 0.9
SOLVER ACCURACY : 1.E-6

Before the simulation destabilizes, there is no other prior time step where the maximum allowable # of iterations is exceeded.

Does anyone have any ideas as to why the simulation might break down so rapidly, and possible reasons for this? One thing to note is that I recently implemented a DRAGFO file to simulate the inclusion of turbine arrays in my domain - it seems the additional shear stress might be causing the problems, but I am already using a very small time step and high solver accuracy. Thanks!

Best,
Alex

Hello Alex

It is hard to guess the cause of exceeding maximum iterations. However, as you have suggested, it is mostly due to the tough forcing of the momentum equation added by drag forces.
Thus try to reduce the time step at least for a short while and then increase it depnding on the solver performances.

With my best regards

Riadh ATA

Hi Riadh,

Thank you for your advice. I reduced the time step, and also increased the values for "implicitation for velocity" and "implicitation for depth" slightly higher than the default values to increase stability.

The simulation runs smoothly now, but I'm just wondering - what do these two variables mean in physical terms? Also, what is the physical meaning of "free surface gradient compatibility"? Thank you again for your help!

Best regards,
Alex

Hello Alex

Implicitation for velocity and depth are pure numerical tools that define the time resolution scheme of the shallow water equations. They give, basically, the weights between the explicit and implicit part of the resolved terms.

The keyword free surface gradient compatibility is used in cases where unstable free surface is observed. It helps for the stabilisation of the solution.

I hope that this helps

with my best regards

Riadh

Hi Riadh,

Yes that is very helpful, thank you!

Regards,
Alex