open TELEMAC-MASCARET The mathematically superior suite of solvers

I am currently running an hydrodynamic simulation of a gulf here in the Philippines. However, the timestep that I need to set must be very small so that the solution won't diverge. That's around 0.002 sec and anything beyond that will make it unstable. I also set that the Courant number must not go beyond 2. I need a solution with a duration of 30 days at least. Thus, it takes a long time to finish the calculation for which I estimated that it will take more than a month to get it done. Any tips how to improve this? Attached is the .slf, .cli and .cas files for your reference.

Hello,

Can you convert your system coordinates from longitude/latitude to a plane projection, like UTM and try it again? Do not forget to change the GEOGRAPHIC SYSTEM = 2 + ZONE NUMBER IN GEOGRAPHIC SYSTEM = 51? I am not so familiar with your area to say if 51 is the correct number, please double check.
TELEMAC-2D or 3D like to work with plane projections. Your time step should be taken bigger.

Moreover, a few advice:
- do not write the keyword STEERING FILE in your steering file to prevent from issues if it is not the same name as the one in which it is written,
- with Finite Elements, I would not use the keyword DESIRED COURANT NUMBER, so I would remove it.

Hope this helps,

Chi-Tuan

Hello again,

Perhaps UTM 52 may be better?

Chi-Tuan

Hello and thank you for your reply.

I did what you suggested and changed the geographic system and the zone number to 2 and 51 respectively (also I tried 52). But when I try to increase the time step value (this time 0.005 sec), the solution still diverges. Do you have any more suggestions? Thank you in advance.

Can you upload your new geometry file?
Perhaps tell us what the minimum size of elements you have in UTM (m) and what the maximum velocities are expected on the area. You should be able to determine a Courant number with a correct time step.

Maybe around a few seconds could be OK, maybe tens if your mesh is coarse.
With the method of characteristics, you can use rather large time steps.

Hope this helps,

Chi-Tuan

I guess I know where my problem lies. The land boundary I uploaded to Blue Kenue wasn't scaled properly. That's why when I check the size of the elements, the values are ridiculously small and makes no sense. I'll correct this thing first and let you know if it will work fine afterward. Thank you very much for your help.