open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

I'm testing my mixed sediment transport model. At first time, I always have very large negative bed evolution (in the picture at one node the negative evolution is about -100m) at some boundary nodes.

Then I checked those points and found the their sand fractions are always above 85%. So they are treated as sand in the mixed model. But unlike the mud has different critical shear stress TOCE_VASE(J) in different layers, the sand only has one critical shear stress for erosion TOCE_SABLE and it's not large enough for the layers under the first layer. So those nodes keep been eroded. Now I use multiple values called TOCE_SABLE_MIX(J) in the subroutine SUSPENSION_FLUX_MIXTE. And it does prevent those nodes from deeper erosion.

I'm not sure if my problem at boundary nodes is realy caused by TOCE_SABLE (Could it be something else? like the way of treating the sediment inflow?). But is it better to have multiple values in TOCE_SABLE?

Kind regards,
Qilong

Hi Qilong,

It looks like an initialization problem. Do you get large erosion along the entire bondary? Does this happen at the first time step?

Could you please check whether this comes from the bedload or from the suspended load. If this is due to bedload, beware there have been a number of changes in 6.2 in the way to impose a bedload transport rate (see enclosed file). If the problem comes from the suspended load, it is sometimes difficult to impose the correct suspended load at the entrance of the domain.

One way to go around this would be to cancel the bed evolutions at the entrance (for a few tens of meters depending on the size of the domain). The only drawback is that the model is no longer mass conservative.

Let me know how this works,

Cath

Hi Cath,

Thanks for your advice!

I checked my results and yes the large erosion occurs along the entire boundary. It doesn't happen at the first time step but the erosion rates at those boundary nodes are quite large. Besides I found there are some nodes not along the boundary (mainly in the shallow areas) also having large bed evolution (large positive (80m) and negative values (-100m)).



I also tried without bedload. The boundary nodes have much smaller erosion but the problem remains at those nodes in the shallow area. So I think the problem isn't caused by the boundary conditions.

And this only happens in v6p3. In the previous v6p2 I never have this problem.

Qilong

Hello,

I'm testing my model with Telemac v6p3 and I have found that my problem comes from the suspension and probably is caused by the way of dealing with the tidal flats. Before I used the OPTION FOR THE TREATMENT OF TIDAL FLATS = 2. Now I'm using OPTION FOR THE TREATMENT OF TIDAL FLATS = 1 and the result becomes better.

Best regards,
Qilong