open TELEMAC-MASCARET The mathematically superior suite of solvers

I want to cut off the sediment supply so that my inlet region scours. (e.g the condition downstream of a dam). How would I do this? I tried setting LIEBOR = 4 but this did not work

how do I specify the amount of sediment entering at the inlet for a bedload transport simulation?

Thanks
Steve

Hello,

via the conlit.f file! In the Sisyphe manual there are some explanations.

Best regards,
Clemens

Hi Clemens,
Yes I tried that by setting the inflow nodes like this:


QBOR%ADR(1)%P%R(334)=0
QBOR%ADR(1)%P%R(468)=0
QBOR%ADR(1)%P%R(466)=0
QBOR%ADR(1)%P%R(463)=0
QBOR%ADR(1)%P%R(458)=0


but this didn't cause any scour at the inlet. Did I miss anything?

Steve

Hi Steve,

Time dependent sediment inflow seems to work in Sisyphe, you have to set
LIQBOR%I(K)=KENT
however I don't know how to get erosion at the inflow boundary.

Best regards,
Clemens

Hi Steve,

make sure that there is no suspended sediment coming into the flow by EQUILIBRIUM INFLOW CONCENTRATION = NO,
if you want to absolutely sure (but this should be default setting) add
CONCENTRATION PER CLASS AT BOUNDARIES = 0;...
code the bedload to be zero as you did, and make sure that LIEBOR=4 in the conlim file for those points.

Michiel

Hi Michiel and Clemens
Thanks for your suggestions
I tried your suggestions but the evolution at the inlet remains fixed at zero....still not getting any scour at the inlet.

this is what I have in my conlit fortran file

LIQBOR%I(334)=KENT
QBOR%ADR(1)%P%R(334)=0

and in the conlim boundary file I have
LIEBOR=4

Do you have any other suggestions?

Thanks
Steve

Hello,

unfortunately no other suggestions how to do that..
but you could extend your model upstream and define a non erdoable area for the inflow region.
Maybe this is also the better solution if you have highly unsteady scour conditions since otherwise possibly you would get instabilities at the inflow boundary.

Best regards,
Clemens

I find it surprising, usually my problem is tryin to prevent erosion or deposition in/near the inflow boundary. Are the flow velocities high enough to trasnport the sediment?

may be able to can get away with prescribing a wall for the sediment (LIEBOR = 2). This will force the sediment influx to 0, but it may cause some instabilities.

Michiel

Hello,

Yes Sisyphe boundary conditions have been designed with the evolution of bottom in mind, which is a practical point of view, whereas on the numerical point of view, the flux of sediment would be easier to deal with.

I you program a little, you can cancel the flux at boundaries in bedload_solvs_fe (case with FINITE VOLUMES=NO). This flux is computed at the beginning and is T8, you can cancel it in the following loop that modifies EBOR by adding a line T8%R(K)=0.D0 after line EBOR%R(K)=... it will then be cancelled for all Dirichlet points.

BUT :

So far what is done in the subroutine positive_depths (that solves Exner equation as a Saint-Venant continuity equation), is that the flux at boundaries is modified at Dirichlet points so that the evolution gives the requested EBOR, this would interfere with what you want. To overcome this you can suppress the treatment of Dirichlet points by changing KDIR into 999 in the call to positive_depths.

I hope it works (not tested...), but if it works we could think of adding this as a new option.

With best regards,

Jean-Michel Hervouet