open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

Im currently trying to simulate the behaviour of salinity stratification in a small basin (40x13x2m) based on the "stratification" example.
My approach to obtain comparable stratification in different basin types is to start with an initial salinity value of 10 g/l and then start the simulation without any in- or outflow.
However i have an initial mass of 8869 g and a final mass of 4037 and my salinity decreases also in the bottom layer (see attached files). Can you help me solving the problem of mass balance and wrong stratification? Which keyword will I have to adapt.

Best regards

Hi there,

As you may know, salinity stratification generally results when you have mixing (or lack thereof) of waters with different salt content (e.g. freshwater overflowing seawater), or where you have large evaporative fluxes. Although I have not seen your boundary conditions file, you appear to be prescribing salinities corresponding to freshwater throughout the water column at both ends of your domain, which is not consistent with what you are trying to simulate (i.e., in order for stratification to develop, you will need saltwater to leave the domain through at least one boundary, which will then have a non-zero concentration somewhere over the depth). Maybe try prescribing 10 g/l at one end of your domain.

Enda

Hi Enda,

I'm not sure we are talking about the same thing.
Through the INITIAL VALUES OF TRACES =10. I set my salinity to 10g/L at every Node.
As you can see from my boundary file, I have two liquid boundaries with prescribed flowrate = 0 m³/s and prescribed tracer values =0 at the first liquid boundary. The second liquid boundary is set for free tracer, however no salt should be able to leave the system flowrate =0 ?

I really have no clue where my salt is going.

Hello,

In your boundary conditions file, the tracer should be given as prescribed (5) as soon as the velocity is 5, so always 5, whatever the boundary, in your case. When the velocity gets out of the domain, the program will itself relax the condition and will change to free exit (4). You should also use the keyword :

TREATMENT OF FLUXES AT THE BOUNDARIES : 2;2

so that the flux of salinity is really .0 if the discharge is zero (otherwise the prescribed value is brutally set, regardless of the flux).

With best regards,

Jean-Michel Hervouet

Hi Jean-Michel,

thank you for your advice. The salt mass now stays the same throughout the whole simulation! I expected that I would see some kind of shift from my completely mixed 10 g/L salt concentration in the basin to some kind of stratification after 5h. However, the basin is still completely mixed in the end.

I understand that diffusion goes in every direction and the density is equal in the whole basin but shouldnt the bottom layer have the highest salt concentration at the end?

Best regards
Simon

Hi Simon,

This was what I tried to explain to you in a previous post. You need to think more carefully about the physical processes that actually cause salinity stratification. Salt dissolves in water, so particles generally do not "settle" (like suspended sediment for example). When you specify an active tracer in TELEMAC, it is a dissolved substance. The only way I can see that you could see a density difference or stratification in your domain is by introducing a fluid with a different salt concentration, or by introducing rain (negative evaporation!) with a different concentration.

See the following link for a description of the difference between solution and suspension - hope this helps:
scienceline.ucsb.edu/getkey.php?key=521

Enda

Hi Enda,

you are right of course, I just anticipated a different behaviour of the telemac salinity library intuitively.
I am now confidently using the proposed two step approach to fill some salt into my basin for one hour and then stop in- and outflow for some time to achieve an even stratification.

Thanks again
Simon