Hej,
I’m working on a lake model, where the water level is fluctuating up to a difference in water level of 2m. The model takes the tributaries and the outflows into consideration.
As the lake itself is rather small (12km2) and also quite shallow, I have to tackle wetting and drying of cells at the boundaries.
To make sure the tributaries and outflows do not run dry, I adjusted the mesh with lowering the bathymetry at this areas. I made sure that there are no steep areas in the mesh.
The current settings for my tidal flats are.
TIDAL FLATS = YES
OPTION FOR THE TREATMENT OF TIDAL FLATS = 1
TREATMENT OF NEGATIVE DEPTHS = 2
MASS-LUMPING ON H = 1
CONTINUITY CORRECTION = YES
SUPG OPTION for water depth = 0
In my liquid boundary file i start at time step 0 with no flow and the next one is 86400 with real measured flow. My TIME STEP in the model is set to 8.
The current model results in large instabilities along the boundaries and the water level is fluctuating a lot (see picture).
How can I increase the stability of the model? Should I adjust the liquid boundary file to allow for a slower “discharge ramping” into the domain?
I already tried a TIME STEP of 2, which did not improve the situation?
Thanks for any help.
Hi
Variations of the water level seen on your graphics are relatively small (lower 1 cm).
I don’t suspect the tidal flats treatment to be the cause of these oscillations. If you really want to be sure, you could try to run the same simulation with a modified bathymetry (lower) the ensure there is no tidal flat…
This could be more linked to the type of boundary condition you have, particularly if you have stage discharge curve downstream…
For me such BC could lead to such small variations linked to the section of the boundary itself…
Hope this helps
1 Like
Thanks for the help.
Yes you are right, they are in the range of about 1cm, which is not that much, but the velocity is also effected by that. Main purpose of the model is to later on look at the shear stresses, which will be a bit complicated if the model is not stable.
The downstream outflow is currently set as a boundary with prescribed flow, where the flow is negative. I added a screenshot of the velocity close to the boundary (outlet), but this behavior is the same along the boundary.
For simplicity I added the BC files. I’m grateful for any comments and suggestions, what could be adjusted to get the model more stable.
Thanks in advance.
VOM_BOUND.cli (49 KB)
Hi
Giving only prescribed flows is the key problem.
Discharge is converted to velocity and this depends on the water level so small fluctuation of water level leads to small variation of the velocity.
Don’t know how this lake operates in reality but this will be probably better to have at least one/few boundary with prescribed level…
Hope this helps
1 Like
Thanks [name removed],
That sounds very reasonable. The lake is regulated but the hydrodynamics of the outflow control unit are not well known. I will test changing to prescribed level one outflow and hope this may make the model more stable.
Supposing it would be a natural lake, where only the discharges but not the levels are known, what would be the most appropriate way to model it?
As an answer to your last question, I would reverse the question to how the discharges are known and not the level.
Measuring a water level is easy and simple contrary to the measure of a discharge.
I don’t have any simple answer to give you a solution to such problems. probably different options should be explore: use sources instead of boundary, modifying locally the geometry of the outlet (lowering local bathymetry to create a pool will reduce velocities) …
Hope this helps
Hi [name removed],
Thanks for your reply.
Could not focus on the model recently due to other obligations.
I have to correct myself, since I had wrong information. The water level is actually known and it seems that the discharge is actually uncertain as the relationship between water level and discharge is not well known.
The discharge from the lake is actually a regulated sluice gate.
In the current solution the bathymetry is lowered at the outlet and the water level was used as boundary. However, after a certain time, when the water level gets “stable” the model still has the same instabilities.
I will post the model files and boundary files later today and hope this clarifies my questions.
Could you please elaborate on the solution using sources, not sure what you mean there.
Thanks so much.
Here are the model files as .zip.
[name removed]
v04.zip (189 KB)
