open TELEMAC-MASCARET The mathematically superior suite of solvers

Dear all

I want to ask before calculate the sediment transport, whether the velocity on all time steps (both steady flow and unsteady flow) needed for the sediment transport model was calculated. If it is ture, than it is not consider the river bed change effects on the flow velocity.

Becuase, In the current SISYPHE user manual, It is said TELEMAC considers the bed change effect on the flow field. it is explicitly. That means, in one time step they assumed that there is no-bedchange during the calculation of flow and then no-flowchange during the calculation of sediment transport/ bed change. After every time step the bed elevation will be updated and used further for the calculation of flow in the next time step.

Because when I run the SISYPHE expmple case (River-arc), I alway found the SISYPHE calculation time is always less than the Telemac2d calculation time.

Regards,

Harry.

Hi Harry,

When running Sisyphe from an existing flow file, it will interpolate between the flow time steps and then correct the flow velocities for the bed changes using the conservation of mass.

Dear Michiel Knaapen,

Thanks for you answer, but I still think SISYPHE didn't consider the bed changes effects on the velocities, because on that case, when you run the SISPHE at least it should coupling the TELEMAC2d/3d model inside on it. Take the River-arc for example, I need run the TELEMAC2d and SISYPHE independently. that means, it can only take the result form TELEMAC2d when we run SISYPHE. how could we take into the effect of the river bed change to the velocity ?

I am not so familiar with those codes, but I am familiar with the back theory on it.

Dear Harry,
Typically, in a loop T2D-Sisyphe what you are solving is:
(i) first the SWE, that is the system containing \partial h/\partial t; \partial hu/\partial t and \partial hv/\partial t
(ii) then, variables (h,u,v) come into sisyphe to solve the sediment transport discharge and the evolution equation \partial b/\partial t

(iii) At this moment, we are not solving the morphodynamics problem as a whole, that is (\partial h/\partial t; \partial hu/\partial t; \partial hv/\partial t and \partial b/\partial t). The implementation of this fully-coupled approach has been discussed but it is not planned for the moment.

I hope it helps,

all the best,

Pablo

Dear Pablo,

Thanks for your helpful answer, Now, I think I know how SISYPHE works. What you said its right. the current version of SISYPHE, it is a uncoupled approach.

I think in the real river, except the pier scour, it is unnecessary to conside the full coupled model. but for the long time river bed change I think it would be better to consider a Semi coupled model. That means, We could set the time steps on SISYPHE much bigger than TELEMAC. and re coupled the TELEMAC 2d again.

I will read the codes on the SISYPHE and check it whether it is possible or not. and we could also discuss this topics. But on codes develops it is seems that this method would be more complicated than full-coupled approach.