open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi,

I have finally succeeded in running a tidal model case with tidal constituents forcing on the boundary, using TPXO derived tidal constituents. (Courtesy Chi-Tuan, www.opentelemac.org/index.php/kunena/16-...using-telemac2d#8685) Currently I am running the model with TELEMAC 3D.

I get satisfactory comparison of predicted currents with observed, when the manning friction = 0.02, VERTICAL TURBULENCE MODEL = 2, and HORIZONTAL TURBULENCE MODEL = 1, .ie, vertical is "mixing length model" and horizontal is "constant viscosity model",

However when I try to set VERTICAL TURBULENCE MODEL = 3, and HORIZONTAL TURBULENCE MODEL = 3 (K-Epsilon model), I get the following error.



I had previously run another model with K-Epsilon model in TELEMAC2d for a smaller domain in the same region of Arabian sea, which has given very close water current calibration, similarly I want to try K-Epsilon model for TELEMAC3D to check the currents calibration. Therefore I wish to run with similar option for TELEMAC3D.

Which other parameters I may have to be change in the steering file, so that the model works with K-Epsilon option.

Can anyone explain me as the significant difference in selecting the different turbulence model type, as in case its implementation, stability etc.

I have also attached the steering file and listing file.



Thanks,
Gourish

Hello,

The only thing I would change in your steering file is :

COEFFICIENT FOR HORIZONTAL DIFFUSION OF VELOCITIES = 1.E-5
COEFFICIENT FOR VERTICAL DIFFUSION OF VELOCITIES = 1.E-5

They both should be set to 1.E-6 because they are taken as molecular viscosity by the k-epsilon model. Otherwise k-epsilon is tricky, especially with tidal flats, so you can try to change EMIN and KMIN in cstkep.f, from 1.D-10 to 1.D-7 or 8 for example. You probably have somewhere huge values of diffusion.

Regards,

JMH